FACULTY OF LAW AND CRIMINOLOGY

Aruna MICHIELS

0509656

Opposition Groups Armed with the

Laws of War and Rational Choice:

Is Compliance the Outcome?

Promotor: Stefaan SMIS

Jury: Jonathan HOLSLAG, Kristof GOMBEER

Academic Year 2015-2016

Master Thesis Submitted in partial fulfilment of the requirements for the diploma Master

of Laws in International and European Law

2

3

Acknowledgements

This dissertation poses the final step in obtaining a Master’s degree in International Law. This

would not have been possible were it not for my parents’ efforts to provide me with the

opportunity of pursuing a higher education, the autonomy to follow my own path, the financial,

logistic and moral support needed and the pedagogical and intellectual stimulation to develop

the necessary skill-set to finally bring this chapter of my life to a conclusion. To the both of you,

I am very grateful and proud to be your daughter. Thanks for believing in me.

Special thanks goes out to my professors Stefaan Smis and Jonathan Holslag, who gave me the

chance to pursue this topic. Thank you for teaching me in your respective fields of expertise,

International Law and International Relations theory. For encouraging my ideas, for the advice,

the feedback and guidance, this dissertation would not be as close to my aspirations had it not

been for you. Thank you for being open and receiving and for communicating with me on the

same level. For steering me in the right direction when the idea for this topic was still a misty

blur in my mind, I would like to thank Kristof Gombeer in particular. His advice got me off to a

good start. To professor Eric David at the Université Libre de Bruxelles I am thankful too for the

stimulating course Droit des Conflits Armés, leading me to explore International Humanitarian

Law further. The European University Institute Library in Florence is worth mentioning too:

thank you for receiving me, for the beautiful collection and the inspiring scenery. To the Vrije

Universiteit Brussel, I am grateful for the existence of this academic institution where I finally

felt to be in the right place.

Finally I would much like to extend my gratitude to friends and family for supporting me. Thank

you, Pieter, for pushing me to set myself to a higher standard. Thank you, Feli, for your

support, for the Photoshop skills and for so much more. For the hospitality and the feedback,

thank you, Manu. For proofreading, thank you, Anke. For the constructive feedback, thank you

Yannick. To my colleague Laurence, thanks for the feedback and sharing the experience. All of

you have greatly contributed to the conception of this dissertation.

4

5

Table of Contents Acknowledgements .......................................................................3

Table of Contents..........................................................................5

Table of Figures ............................................................................9

1 Introduction ............................................................................ 11

1.1 A GENERAL LACK OF COMPLIANCE..................................................11

1.2 METHODOLOGY ............................................................................14

2 Definitions & Point of Departure ................................................. 15

2.1 THE CONCEPT OF COMPLIANCE ......................................................152.2 THE NOTION OF NORMS ................................................................17

2.3 CHALLENGES FOR IHL COMPLIANCE................................................18

3 A Rational Choice Paradigm for IHL Compliance............................ 23

3.1 A RATIONALIST-INSTITUTIONALIST APPROACH................................233.2 UNDERLYING ASSUMPTIONS..........................................................26

3.3 A RATIONAL CHOICE APPROACH TO IHL OBLIGATIONS .....................353.3.1 Conventional sources ..........................................................................353.3.2 Customary sources .............................................................................383.3.3 Illustration of the dynamic...................................................................413.3.4 Preliminary conclusions .......................................................................43

3.4 ENFORCEMENT AND SUGGESTIONS FOR INCREASING COMPLIANCE ...443.4.1 Enforcement ......................................................................................443.4.2 Possible areas of improvement .............................................................45

4 Armed Opposition Groups and IHL Compliance ............................. 49

4.1 EXTRAPOLATING THE PARADIGM TO NON-STATE ACTORS .................494.1.1 Delimiting the scope ...........................................................................494.1.2 The scope of application of IHL ratione materiae et personae ...................504.1.3 Do the assumptions hold up? ...............................................................58

6

4.2 APPLYING THE RATIONAL CHOICE MODEL FOR IHL COMPLIANCE TO

AOG’S ..............................................................................................614.2.1 Conventional sources ..........................................................................624.2.2 Customary sources .............................................................................634.2.3 Pouring the info into illustrative models .................................................644.2.4 Practical implications ..........................................................................68

4.3 ENFORCEMENT AND SUGGESTIONS ................................................704.3.1 Enforcement ......................................................................................704.3.2 Challenges and possible areas of improvement.......................................70

5 Conclusion .............................................................................. 74

5.1 EVALUATIVE REMARKS..................................................................74

5.2 CONCLUSION...............................................................................75

Bibliography............................................................................... 78

LEGISLATION ....................................................................................78

DOCTRINAL SOURCES ........................................................................79Books .......................................................................................................79Contributions in Collective Doctrinal Works ....................................................83Articles .....................................................................................................87Other........................................................................................................89

Appendix ................................................................................... 91

SCHEMATIC OVERVIEW OF IHL SCOPE OF APPLICATION .........................91

DETAILED JUSTIFICATION OF TABLES AND MODELS INCLUDED ...............92Table 1: A simple coordination game.............................................................92Table 2: A one-shot Prisoner’s Dilemma.........................................................92Table 3: Assurance game: trustworthy players ...............................................93Model 1: Tree model for a sequential game....................................................93Table 4: Payoffs for unconstrained states.......................................................94Model 2: Sequential game for unconstrained states.........................................95Table 5: Payoffs for constrained states ..........................................................95Model 3: Sequential game for constrained states ............................................96Table 6: Payoffs for unconstrained insurgent ..................................................96Model 4: Sequential game for unconstrained insurgent ....................................97Table 7: Payoffs for constrained insurgent .....................................................97

7

Model 5: Sequential game for constrained insurgent........................................98Table 8: Payoffs for unconstrained belligerent ................................................99Model 6: Sequential game for unconstrained belligerent...................................99Table 9: Payoffs for constrained belligerent ..................................................100Model 7: Sequential game for constrained belligerent ....................................100

8

9

Table of Figures Table 1: A simple coordination game .............................................................................. 29Table 2: A one-shot Prisoner's Dilemma .......................................................................... 29Table 3: Assurance Games: trustworthy players ............................................................... 30Model 1: Tree model for a sequential game ..................................................................... 33Table 4: Payoffs for unconstrained states ........................................................................ 42Model 2: Sequential game for unconstrained states .......................................................... 42Table 5: Payoffs for constrained states............................................................................ 42Model 3: Sequential game for constrained states .............................................................. 43Table 6: Payoffs for unconstrained insurgent.................................................................... 64Model 4: Sequential game for unconstrained insurgent...................................................... 65Table 7: Payoffs for constrained insurgent ....................................................................... 65Model 5: Sequential game for constrained insurgent ......................................................... 66Table 8: Payoffs for unconstrained belligerent .................................................................. 66Model 6: Sequential game for unconstrained belligerent .................................................... 67Table 9: Payoffs for constrained belligerent...................................................................... 67Model 7: Sequential game for constrained belligerent........................................................ 68

10

11

Le droit des conflits armés est probablement la branche la

moins respectée et, par conséquent, aussi la plus

théorique, sinon la plus utopique du droit international, et

même du droit tout court!1 – Eric David

1 Introduction

1.1 A GENERAL LACK OF COMPLIANCE

The title of this dissertation tries to abstractly grasp all of its key features: the Laws of War will

be approached through Rational Choice Theory in order to draw conclusions on compliance

behaviour of Armed Opposition Groups. As one might derive from the quote above, compliance

with humanitarian standards is no given. The Laws of War are still much needed in our

contemporary society, which is no stranger to armed conflict. The body of International

Humanitarian Law (IHL) saw the light of day in the context of warring Europe in the 19th

century.2 It seeks to alleviate some of the horrifying consequences of warfare by creating

standards for protection and to regulate what means are acceptable for modern warfare. Some

are willing to endorse IHL norms publicly, yet their actions are testimony to a lack of

commitment in practice.3 Even in the face of widespread ratification of part of the instruments,

violations occur frequently.4 Moreover a substantial part of the standards by now has obtained

the status of Customary International Law, binding actors that have not expressly consented to

these obligations.5 Despite proliferation and expansion of the IHL normative framework,

effective implementation and enforcement of IHL are hard to accomplish. A lack of compliance

with the legal obligations it engenders has been amply established. The ICRC has pointed out

1 E. DAVID, Principes de Droit des Conflits Armés, Brussels, Bruylant, 2012, 695; (Hereafter E. DAVID,

Principes). 2 C. ZOLI, E. SCHNEIDER & C. SCHUSTER, “Armed conflict and Compliance in Muslim States, 1947-2014:

Does Conflict Look Different under International Humanitarian Law?”, N.C.J. Int’l L. & Com. Reg. 2015, 694. 3 P. MAURER, Strengthening Compliance with International Humanitarian Law, Speech at the 4th meeting of

States on Strengthening Compliance with Humanitarian Law, Geneva, 23 April 2015,

https://www.icrc.org/en/document/fourth-meeting-states-strengthening-compliance-international-

humanitarian-law. 4 The Geneva Conventions are accepted to be universally ratified, f. e. - Resolution 1 – Strengthening legal

protection for victims of armed conflict, 31st international Conference 2011, ICRC, 1 December 2011; T.

MERON, “On the Inadequate Reach of Humanitarian and Human Rights Law and the Need for a New

Instrument”, AJIL 1983, 590; J. M. HENCKAERTS, “Study on customary international humanitarian law: A

contribution to the understanding and respect for the rule of law in armed conflict”, IRRC 2005, 187. 5 T. MERON, “The Humanization of Humanitarian Law”, AJIL 2000, 242-244.

12

that noncompliance is the number one cause of civilian suffering in armed conflict.6 In fact, no

less than 90% of armed conflict death toll consists of civilian deaths, mocked up to collateral

damage.7 These figures are probative of the undeniably urgent nature of this issue.8 An ever-

growing body of literature further supports its topical relevance. As the 90’s ran its course

scholarly doctrine took an interest in International Law Compliance.

This trend was paired with the simultaneously emerging trend amongst academia to address

this issue from an interdisciplinary point of view, combining International Law (IL) with

International Relations Theory (IRT) to capitalise on the interplay between law and politics. It

has been said that compliance has become a central preoccupation in international legal

scholarship.9 This spiked my interest to further explore the subject. Even though both disciplines

are valuable on their own merits and approach IL differently, there is a significant contentious

overlap. It is my strong opinion that it is important to bring IL and IRT closer together. I sense

upholding this dichotomy is unnatural and serves little purpose. Along with my predilection for

IHL since taking a course with Prof. Eric David at the ULB, this led to the idea to write this

dissertation on IHL compliance.

Another observation significantly impacted the basic idea: the context in which IHL operates

nowadays has undergone substantial changes. Whereas the majority of armed conflicts were

inter-state wars at the time IHL was developed, now most hostilities cannot be categorised as

such. The greater part is qualified as Non-International Armed Conflicts;10 some occurrences of

political violence do not meet the criteria of armed conflict;11 the methods and means of warfare

mark astonishing evolutions of modern warfare.12 The compliance literature has converged on

6 Resolution 1 – Strengthening Legal Protection for Victims of Armed Conflict, 31st international Conference

2011, ICRC, 1 December 2011. 7 N. MELZER, “Bolstering the Protection of Civilians in Armed Conflict”, in A. CASSESE (Ed.), Realizing

Utopia: The Future of International Law, Oxford, Oxford University Press, 2012, 509; (Hereafter A.

CASSESE (Ed.), Realizing Utopia). 8 J. HERMAN, “International Law and humanitarian space in the twenty-first century: challenged

relationships”, in A. ZWITTER, CH. K. LAMONT, H.-J. HEINTZE & J. HERMAN (Eds.), Humanitarian Action:

Global, Regional and Domestic Legal Responses, Cambridge, Cambridge University Press, 2015, 12-13. 9 S. MENDLOVITZ, “Unrealistic Compliance Goals”, ASIL Proceedings 1964, 9. 10 D. THÜRER, “International Law: Theory, Practice, Context”, in Collected Courses of the Hague Academy of

International Law, The Hague, Hague Academy of International Law, 2011, 199; S. SIVAKUMARAN, “Binding

Armed Opposition Groups”, International and Comparative Law Quarterly 2006, 396. 11 Though they do exceed the level of violence regular law enforcement can cope with.- See N. MELZER,

“Bolstering the Protection of Civilians in Armed Conflict”, in A. CASSESE (Ed.), Realizing Utopia, 510-511. 12 New weapons technologies such as targeted killings, drone warfare, cyber warfare and more have

changed the face of IHL. For a more detailed assessment of the aspect of the implications of weapon

technology under international law, see N. RONZITTI, “Modern Means of Warfare: The Need to Rely upon

13

another new element of the armed conflict context.13 Non-State Actors (NSA’s) are certainly not

a novelty in the international order, yet the intuition that something has changed remains. On

the one hand, as International Organisations evolve, so does their involvement in the

international scene. Third party intervention has manifested for instance through UN mandated

Peacekeeping. Although in this scenario objectives are impartial to the cause of the conflict,

things can still go wrong: even peacekeepers can violate rules of IHL.14 Other causes vary from

outsourcing to Private Military and Security Corporations (PMSC’s), the more frequent

involvement of Non-Governmental Organisations (NGO’s) with a humanitarian agenda, the rise

of Armed Opposition Groups (AOG’s) and civilian involvement in armed conflict.15 This

observation is especially problematic for greater IHL compliance, since a lot of duties and

obligations under IHL are treaty based. For states and International Organisations it’s obvious

they have the legal personality required to enter into treaties. The other actors involved in

armed conflicts might not possess this legal personality per se. The literature offers several

explanations to argue that non-state belligerent actors are bound by IHL, but this evolution

poses another challenge for IHL compliance in fact.16 Moreover these contextual evolutions lead

to a blurring of the standards and norms in place, making it harder to uphold even the most

cardinal principles of IHL like distinction between civilians and combatants. Do the Laws of War

effectively constrain behaviour of actors involved in armed conflict? Can a Rational Choice

analysis of their incentives to comply contribute to improving IHL compliance? These questions

related to states and AOG’s in particular mark the point of departure for this dissertation.

International Humanitarian Law, Disarmament, and Non-Proliferation Law to Achieve a Decent Regulation of

Weapons”, in A. CASSESE (Ed.), Realizing Utopia, 553-570; For an analysis of the legality of targeted

killings under IHL see R. OTTO, Targeted Killings and International Law: With Special Regard to Human

Rights and International Humanitarian Law, Heidelberg, Springer, 2011, 203–350; See also M. A. NEWTON,

“Flying into the Future: Drone Warfare and the Changing Face of Humanitarian Law”, Keynote Address at

the 2010 Sutton Colloquium, Denv. J. Intl. L Pol’y 2010, 602; J. T. G. KELSEY, “Hacking into International

Humanitarian Law: Principles of Distinction and Neutrality in the Age of Cyber Warfare”, Mich. L. Review

2008, 1429; D. SAXON, “Introduction: International Humanitarian Law and the Changing Technology of

War”, in D. SAXON (Ed.), “International Humanitarian Law and the Changing Technology of War”, Leiden,

Martinus Nijhoff, 2013, 1-16. 13 R. KEOHANE, “How Actors in World Politics Behave: Legalizations and World Politics”, in L. HENKIN How

are nations behaving?, ASIL Proceedings 2002, 213. 14 See K. OKIMOTO, “Violations of International Humanitarian Law by United Nations Forces and their Legal

Consequences”, Yearbook of International Humanitarian Law 2003, 199-236; For the ICRC’s point of view

see also: https://www.icrc.org/eng/resources/documents/misc/57jmbh.htm; 15 D. THÜRER, “International Law: Theory, Practice, Context” in Collected Courses of the Hague Academy of

International Law, The Hague, Hague Academy of International Law, 2011, 198-199; See also note 103

infra. 16 N. LUBELL, Extraterritorial Use of Force Against Non-State Actors, Oxford, Oxford University Press, 2010,

95-96.

14

1.2 METHODOLOGY

The above stated changes have resulted in an obvious asymmetry between state versus other

actors involved in armed conflicts, due to an apparent gap in firing power, privileges or

obligations. This lack of reciprocity seems to result in a lack of incentive for non-state

belligerents to comply with the demands of IHL.17 The aim is to combine two international

academic disciplines, International Law and International Relations Theory to cross-fertilise both

with their respective insights. More concretely I will attempt to develop a theoretical paradigm

for compliance with IHL by taking a rationalist-institutionalist approach. Hence, I will try to draw

from scholarship on Rational Choice Approaches for International Law in general and transpose

the paradigm to IHL at the state level. In a next phase I will attempt to apply the resulting

paradigm to NSA’s, in particular to AOG’s in order to investigate if and how their compliance

behaviour can be affected with IHL norms. Rational Choice Theory seems an appropriate choice

of approach, since the problem is characterised by two dimensions: a behavioural dimension

(compliance vs. defection) and a legal dimension: (upholding international legal standards). RCT

departs from a number of assumptions that allow predicting actors’ strategic behavioural

decisions. This approach uses game-theoretic models to objectify actors’ rational calculations of

possible outcomes for compliance and defection. I am aware of the fact this mathematical

approach can only result in highly theoretical findings and there is a considerable gap between

theory on regulating armed conflict and the practical face of battlefield conduct. This approach

is based on the intuition that RCT has a significant potential for IHL compliance through analysis

of actors’ incentives to comply with IHL, violate it, reciprocate or retaliate in armed conflict

situations. The merits or disadvantages of theoretical research are however not under debate

here. I also realise that each discipline has its own language and combining them might cause

difficulties. I will do my best to provide sufficient explanation for any reader to catch the drift of

both. This dissertation does not pretend to have a final say on anything. On the contrary, since

it will not be possible to exhaustively research this topic, its aim is to explore what might lie

around the corner of this path.

The next chapter will address some terminological issues of clarification and is dedicated to

situating the issue areas. In the third chapter I will develop a Rational Choice framework for IHL

compliance on the state level. This will then be extrapolated to AOG’s in the fourth chapter. The

fifth chapter is devoted to realigning what findings surfaced and, to some extent, an evaluation

of the approach taken. In doing so I hope to advocate that interdisciplinary scholarship in

International Law deserves a chance.

17 N. MELZER, “Bolstering the Protection of Civilians in Armed Conflict” in A. CASSESE (Ed.), Realizing

Utopia, 509.

15

2 Definitions & Point of Departure “com-pli-ance / k∂m’plaI∂ns/ n [U] formal

when someone obeys a rule, agreement, or

demand”18 – Longman Dictionary

Before fully embarking on this endeavour, it seems appropriate to get some terminology out of

the way for clarity’s sake. When looking up the term “compliance” in a dictionary, the result

seems simple enough, as the quote above shows. In the scholarly context, terminology is

important to reflect on in order to avoid misinterpretation. The notions of compliance and norms

referred to will be defined and distinguished from other related terms in the following

paragraphs. Then, permit me to elaborate briefly on why IHL seems to be violated more than

other branches of international law before jumping into the compliance theory headfirst.

2.1 THE CONCEPT OF COMPLIANCE

In this doctrinal field compliance is usually defined as “a state of conformity or identity between

an actor’s behaviour and a specified rule”.19 This definition entails the nuance of accordance of

actual behaviour with prescribed behaviour, as was iterated by Oran Young.20 It is to be

distinguished from related concepts of implementation, effectiveness, commitment, obedience

and adherence. In this context implementation refers to the measures, actions or behaviour

states adopt to convert international norms into the domestic legal order, allowing policy to

reflect the norm’s content. Often compliance behaviour requires measures of implementation,

but they remain two distinct concepts nonetheless.21 One might state that compliance goes

further than implementation.

18 P. PROCTER (Ed.), Longman Dictionary of Contemporary English, Harlow, Longman, 1978; online version

at www.longmandictionariesonline.com, v° “compliance”. 19 K. RAUSTIALA & A.-M. SLAUGHTER, “International Law, International Relations and Compliance”, in W.

CARLSNAES, T. RISSE & B. A. SIMMONS (Eds.), Handbook of International Relations, London, Sage

Publications, 2002, 539. 20 M. BURGSTALLER, Theories of Compliance with International Law, Leiden, Martinus Nijhoff, 2005, 5

(Hereafter M. BURGSTALLER, Theories of Compliance). 21 ibid., 4-5; K. RAUSTIALA & A.-M. SLAUGHTER, “International Law, International Relations and

Compliance”, in W. CARLSNAES et al. (Eds.), Handbook of International Relations, London, Sage

Publications, 2002, 539.

16

Effectiveness expresses a degree of change, improvement or realisation as a result of a given

rule, but not necessarily in relation to compliance.22 Compliance behaviour is usually an

indicator of the effectiveness of a rule. This is a logical corollary of the fact that rules are

intended to reach an objective. Where compliance is linked with behaviour, effectiveness is not;

it is merely an autonomous notion reflecting the extent to which the objectives of the rule are

attained.

Commitment implies the actors’ agreement with a norm or their decision to agree. Often this

also implies their willingness to comply. Committing to a certain course of action is also meant

to convince others the actor will act accordingly.23 This implies commitment precedes compliant

behaviour. Behaviour establishing non-compliance would not undermine the previous act of

committing to a norm. The intention to comply usually does incite commitment, but does not

presuppose it, nor require it.24 This notion can be associated with the concept of legal obligation,

which flows from a formal commitment. Consent-based theories are centred around the idea

that international law is only a source of legal obligation insofar actors commit to it by

expressing their consent. Legal obligation is an external factor for compliance, since the actor is

theoretically obliged to adapt its behaviour in line with the rule. Yet, this view does not take into

account the innate reasons or incentives to comply. The notion of compliance that underpins

this dissertation goes further as it will probe why actors comply. Two reasons for this are that

not all legal obligations entail compliant behaviour (hypothesis of violation of rules committed

to) and not all actors involved are under legal obligation but still portray behaviour that is either

compliant or not (hypothesis of compliance to a rule not formally committed to).

The difference with obedience in the literature refers to a process of internalization of a norm,

coined in Koh’s transnational legal process approach.25 Obedience appears as a voluntary or

automated form of behaviour, whereas compliance appears as coerced by the external norm.

One might reformulate this as the distinction between internal coercion, where the need to

comply appears the result of an innate process, versus external coercion where compliance is

demanded from forces alien to the actor. From my point of view both these dimensions

correlate, as behaviour is always the result of internal and external variables interacting.

22 M. BURGSTALLER, Theories of Compliance, 4. 23 K. T. GAUBATZ, “Democratic States and Commitment in International Relations”, in B. A. SIMMONS & R.

H. STEINBERG (Eds.), International Law and International Relations, Cambridge, Cambridge University

Press, 2006, 45. 24 Burgstaller posits that the commitment in itself does not provide states with a sufficient incentive

structure to comply with a given rule. Compliance implies a more profound incentive to see the rules’

objectives realised, regardless of formal commitment. - M. BURGSTALLER, Theories of Compliance, 3-4. 25 F. B. ADAMSON & C. L. SRIRAM, “Perspectives on international law in international relations”, in B. ÇALI

(Ed.), International Law For International Relations, New York, Oxford University Press, 2010, 26-27; é

17

Adherence is distinguished from compliance through a temporal criterion. Whereas adherence is

limited to the actions taken by the actor immediately after he is bound by this rule, compliance

implies further action in line with said rule as time elapses. This is relevant in the context of

international law and policy, as public officials succeed one another. More concretely,

governments competent to shape domestic and foreign policy can (choose to) be bound by a

certain international rule and adhere to it. To give effect to this rule implies the enactment of it

as a national rule. This is usually followed by an international decision to engage in concerted

action. A subsequent government then finds itself in a situation that might have undergone

societal or contextual changes that influence the policy it has to adopt at a later point in time.

At this point two options arise: the subsequent government can decide to comply with the

previously adhered to rule or defect from it. From this distinction it becomes apparent that

compliance implies the continuation on a chosen path as a result of a conscientious decision. 26

Summarizing, one might say that the notion of compliance covers certain specific nuances that

become clear when distinguishing it from related notions. It appears to be behaviour in line with

a given rule, resulting from external coercive factors, portrayed over time. Compliance can lead

to enhanced effectiveness of the rule complied to. Implementation of said rule can illustrate

compliant behaviour. The notion of compliance has now been adequately delineated. In turn,

the notion of norms will be offered the same courtesy in the next paragraphs.

2.2 THE NOTION OF NORMS

The notion of norms used in the sphere of international law and relations varies in each

discipline. International Relations literature uses a fairly broad notion of norms, including but

not limited to legal rules. There is an apparent tendency in IRT literature to avoid referring to

laws or legal rules when circumscribing the notion of norms.27 The notion this dissertation will

apply refers to norms of IHL, thus essentially legal rules. A generally accepted notion of norms

defines them as standards of appropriate behaviour for actors.28 These standards are usually

construed in terms of rights and duties. A legal rule entails a more nuanced definition,

distinguishing it from the broader termed norm. Rules are different as they prescribe or

26 J. TRACHTMAN, “International law and Domestic Political Coalitions: The Grand Theory of Compliance with

International Law”, Chicago Journal of International Law 2010, 129-139. 27 Harold Koh goes so far as referring to law as the “L-word”. See H. KOH, “Why Do Nations Obey

International Law?”, The Yale Law Journal, 1997, 2625. 28 M. BURGSTALLER, Theories of Compliance, 81.

18

proscribe specific behaviour.29 Legal rules have varying natures and forms. In the context of this

subject they mainly pertain to conventional and customary sources. IHL used to be an

essentially customary branch of IL, but has evolved due to codification in multilateral treaties.

Nevertheless, customary international humanitarian law still completes and adapts IHL, since

the conventional sources are often outdated in light of current practice.30

Obviously, law is undergirded with general principles, culture, societal values etc.: the material

sources of law. Those are considered to be presupposed and an integral part of the rule. It does

not seem useful from a methodical point of view to distinguish these for the purpose of this

theoretical analysis. This dissertation aims to research IHL compliance in general and not in a

certain society, time span or geographical space. Doing so, I assume the international nature of

the legal body generates a globally agreed upon content. Obviously this does not correspond

with reality as some aspects discussed might illustrate, but for the sake of simplicity and

generality, permit me as much.

Finally, as the nuance of behavioural standards is of special import for the compliance literature,

it will often seem more appropriate or justified to use the term ‘norm’ or ‘standard’ instead of

‘rule’. This has everything to do with the key role of incentives and their internal or innate

nature in determining actors’ behaviour. The notion of ‘norm’ will thus refer to an internalised

standard. The notion of rule will refer to the external body of obligations the actor is subjected

to. The term standard will apply to both as a general umbrella term. The object of all three

terms is legal in nature. As I have clarified some essential terms by now, the following

paragraphs are devoted to shedding some light on the issue area this dissertation aims to

investigate. This will serve as a point of departure for the analysis to follow and to situate the

problem of noncompliance in IHL.

2.3 CHALLENGES FOR IHL COMPLIANCE

The question of compliance begs off course the question of violation. The body of IHL operates

in a context where state security and survival are at stake. These are high stakes. If all parties

to the conflict are willing to comply, this is obviously a mutual best outcome, since on both

sides this would amount to considerably less damage and deaths. However, if either of the

parties involved decide not to do so, they gain a considerable military advantage and increase 29 S. D. KRASNER, “Structural Causes and Regime Consequences: Regimes as intervening Variables” in B. A.

SIMMONS et al. (Eds.), International Law and International Relations, Cambridge, Cambridge University

Press, 2006, 3. 30 E. DAVID, Principes, 61; 71. See infra 35, no. 3.3 .

19

their potential to secure victory. The fact that IHL operates in such a high-stakes environment

engenders certain implications in the theoretical model of IL compliance this dissertation seeks

to apply. I will elaborate on this further under chapter three where the theoretical model will be

developed. The above represents a first problem area for IHL compliance.

The agential gap between actors committing to IHL obligations and the actors that need to

observe them in practice generates a second problem area for IHL compliance. The states’

soldiers are the ones making the actual behavioural decisions to comply during combat. This

results in several challenges on each level as James. D. Morrow has accurately described in his

recent work Order within Anarchy.31 He distinguished three sets of problems on each level that

can lead to escalation of violence, often in its most extensive and deplorable form. At the state

level IHL is impeded when violations are committed as the result of deliberate state policy not

to comply (1). Such a decision reflects either a strategic calculation of military advantage, or a

fundamental rejection of the standards in place. A very recent example of such violations is the

use of chemical weapons by the Assad regime in Syria.32 Such behaviour signals to the opposing

parties a resolute lack of commitment, triggering reciprocal noncompliance or even retaliation.

Opportunistic defection (2) poses another challenge for IHL. The difference being that states are

more likely succumb to the temptation of securing the advantage ad hoc rather than

determinately flouting the rules as a matter of policy. A third scenario is inherent to the nature

of international law: self-interested interpretation (3). Rules can rarely regulate situations

exhaustively and will always leave some room for interpretation when straightforward

application fails. In doing so, subjects of law are tempted to make efforts to justify actions in a

legal grey zone.

At the individual level, two more causal factors generate challenges for IHL compliance: a

problem of agency, or how states exert effective control over its agent-soldiers to induce

compliance and the problem of noise, or the difficulty of judging the other side’s intentions of

(non) compliance. For one, inadvertent violations (1) might occur due to the uncertainty of

identification and the inaccuracy of weapons33. Second, perfidy (2) –the use of IHL protection in

31 J. D. MORROW, Order within Anarchy: The Laws of War as an International Institution, New York,

Cambridge University Press, 2014, 63-70. (Hereafter J. MORROW, Order within Anarchy). 32 Resolution 2209 of the United Nations Security Council (6 March 2015), UN Doc. S/RES/2209 (2015);

Report of the Independent International Commission of Inquiry on the Syrian Arab Republic, Human Rights

Council, A/HRC/23/58, 4 June 2013; T. RODENHÄUSER, “International legal obligations of armed opposition

groups in Syria”, International Review of Law 2015, 16 p. accessed online at

http://dx.doi.org/10.5339/irl.2015.2.

33 The accuracy of weaponry has obviously increased significantly due to technological progress (drone

warfare, guided weapons systems, targeted killings for example). Though these evolutions allow for more

20

bad faith to gain an unfair advantage- poses another challenge. The fundamental principle of

distinction is abused in order to protect targets otherwise legitimate. Actions of the sort

undermine the credibility of mutual compliance and results in more noise, impeding correct

situational assessments of observed behaviour. A third problem lies where individual

combatants decide to violate the laws of war in contravention of ordered policy (3). The military

structures nowadays have built-in mechanisms to bring troops to order, discipline them or

adjudicate them for misconduct. Yet, the distance in the chain of command between the states,

the high-ranking officers and the lower ranking soldiers can allow for noncompliant behaviour

eluding oversight. One can imagine officers preferring to turn a blind eye to selling out their

men who are merely doing the dirty job in the chaos of armed conflict.

This third problem ties in neatly with another strenuous aspect of IHL violations. Since IHL

violations are always actions or behaviour of people, and a lot of different variables of different

natures (legal, ideological, psychological, geopolitical etc.) determine behaviour, it is no easy

task to pinpoint exactly what causes IHL violations to occur from a psycho-sociological point of

view. Since violations of IHL can only occur in the context of armed conflict, the causes

underlying the conflict are an indirect source of defective behaviour. This context of armed

conflict implies exceptional circumstances, governed by different rules deviating from the

normal situation. Eric David distinguishes some causes on two levels of a macro- and a micro-

sociological nature.34

From a macro-sociological point of view he indicates that like any behaviour, violations of IHL

are the ‘product of history’ and the ‘result of an actual situation’. What he refers to, as the

‘product of history’ is the underlying social dynamics of the opposing parties that over time led

them to an armed conflict. The more heated the argument, the greater the temptation to

impose greater might on the enemy, and the closer they get to the brink of violation. In a lot of

cases extreme violations are born from deep seeded quarrel and dissension amongst parties

competing in some societal way. The intensity of the dispute has a report with the frequency of

violation. The frequency of violent conflict is inversely proportional to the amount and might of

“the other side”. Interstate conflict as such appears to be the most intense category and the

least frequent.

The ‘result of a situation’ refers to the actual context of the armed conflict. On the one hand, an

armed conflict provides a context that per definition facilitates or justifies behaviour that would

control over which objectives to target, the collateral damage does not necessarily decrease accordingly.

This debate centres on weighing the principles of military necessity versus proportionality demands against

one another. – see supra 16, note 24. 34 E. DAVID, Principes, 975-1046.

21

be unacceptable during peacetime. The lawful combatant is entitled to perform actions that

under normal circumstances would constitute criminal offenses – in an armed conflict “murder”

translates to “hostile target neutralised” for example. Moreover, the context facilitates illicit

behaviour in three more ways. For one, violence leads to more violence; it functions like a

contagion. Actors exposed to a new standard of violence will mimic violent behaviour. Second,

in the context of an armed conflict the line drawn between acceptable and inappropriate or illicit

behaviour becomes very thin. Actors find it increasingly hard to hold on to values or distinguish

between “good” and “bad” when confronted with danger and destruction. Survival instinct,

mission objectives, fear and exposure to gruesome sights will cause desensitization and dilution

of moral principles. Third, the context of armed conflict is prone to induce psycho-pathologic

conduct and bestiality, such as morbid behaviour, sadism and vengeance. The items in the list

of war crimes of article 8 of the Rome Statute such as pillage, torture and rape amongst many

other cruel behaviours, are not picked at random.

On the other hand, armed conflicts often arise in socio-political contexts founded on unethical or

amoral principles, like racism, xenophobia or exclusion policies, where the opposing party is

dehumanised. Violent behaviour directed toward the “enemy” is encouraged or taught through

several processes. Violent behaviour can be stimulated in several ways. First, indoctrination by

twisted principles is the result of several forces at work: active convincing by the leadership,

passive toleration by third persons and justification or self-persuasion by the actor. Participation

in armed conflict dynamics can start to escalate progressively over time. Growing accustomed

to a level of violence can increase the intensity of violence. Division of labour or tasks that all

together amount to IHL violations, allows for actors participating in one part of the chain to

forget, overlook or justify their part of responsibility for the crime committed in the end. This

often occurs in context where mass murder is structurally organised.

All of the above macro sociological dynamics result in individual actors resorting to IHL

violations in the context of armed conflict. David indicates two more aspects that induce IHL

violations. On the one hand he distinguishes causes in the agent’s personal experiences that

prepared him for IHL violations, in particular his formation, a general or a specific education.

Armed Opposition Groups often implement such strategies for recruitment. Current events can

support such contentions in relation to ISIS, which recruits nearby as well as abroad.35 One can

be taught to obey and follow orders strictly, to contempt the enemy or to be the product of a

culture of victimisation. On the other hand he indicates justifications or rationalizations the actor

35 K. LEGGIERO, “Countering ISIS Recruitment in Western Nations”, Journal of Political Risk 2015, accessed

online at: http://www.jpolrisk.com/countering-western-recruitment-of-isis-fighters/; E. GRAHAM-

HARRISON, “Taliban fears over young recruits attracted to Isis in Afghanistan”, The Guardian 2015,

accessed online at: http://www.theguardian.com/world/2015/may/07/taliban-young-recruits-isis-

afghanistan-jihadis-islamic-state.

22

applies to his misconduct. These rationalizations can be of an ideological, strategic or

economical nature. Ideological justifications can pertain to a sense of patriotism or nationalism,

racist or opposing philosophy or ideology or social conformism. Strategic rationalizations aim at

defeating the enemy or pre-emptive defence. Since violations of rules are, like the rules, man-

made, any improvement would imply to teach the actors involved in armed conflict about

acceptable behaviour and all its aspects –legal, sociological, psychological etc.- in order to

prevent individuals from committing crimes.

These problems can only be addressed in two ways: on the state level by achieving more

widespread endorsement and ratification of IHL and on the individual level, agential problems

require better formation and discipline. All these dynamics characteristic of armed conflict

complicate reciprocal enforcement of the obligations. States and soldiers are required to deduce

whether the other side is willing to uphold the protective standards, judge intentions and

whether violations are deliberate. For further theoretical analysis of IHL compliance, the above

described dynamic is particularly interesting to explore through the lens of rational choice

theory: actors with opposing interests are confronted and required to make decisions to comply

or not to comply. They interpret behavioural signals and subsequently develop strategic

responses.

This invites me to explore what potential RCT has to offer on these matters; it is the reason why

the next chapter embarks on an attempt to formulate a rational choice theoretical model for IHL

for state compliance. A first part is dedicated to situating the interdisciplinary nature of the

approach. The second subchapter elaborates the rational choice assumptions at the basis of the

paradigm. A third part applies the paradigm to conventional and customary sources of IHL. The

fourth part elaborates on the aspect of enforcement and formulates some suggestions to

ameliorate compliance on the state level

23

3 A Rational Choice Paradigm for

IHL Compliance “Silent enim leges inter arma” – Cicero, Pro Milone

3.1 A RATIONALIST-INSTITUTIONALIST APPROACH

IHL is known to be paradoxical in nature, providing codes of conduct for the worst-case scenario

of war, where no rules of conduct seem to hold up. Cicero duly pointed out the normative void

war bears in its nature in his address to Milone. Yet during the nineteenth century the decision-

makers felt the need to regulate conduct in conflict. A body of treaty law binding states is the

result. This chapter will focus solely on states as unitary actors making decisions as the product

of policy in relation to humanitarian obligations. The legal obligations rest primarily on states as

they are considered unitary actors endowed with International Legal Personality (ILP), required

for entering into treaties. This fictitious personification allows state representatives to act for the

collective entity of national subjects.36 Most branches of IL are thought to regulate the relations

between states only. IHL deviates from this approach by applying an actor-centric approach,

regulating conduct for all group or individual actors involved in armed conflicts. 37 The extent to

which NSA’s are bound by IHL will be discussed under chapter four.

In the IRT literature several currents of thought offer different views on IL and why states enter

into treaties. The model I will elaborate on is based on the rationalist-institutionalist train of

thought. I will shortly present the main arguments of three important strands of IRT -realism,

liberalism and constructivism, since institutionalism carries features from both realist and

liberalist paradigms and contrasts with constructivism.

First, the realist approach views IL as epiphenomenal to state power and denies IL can

independently affect state behaviour. States will only comply with IL as long as it serves

national interest and state security. They view IHL as the by-product of state interests and

influence of powerful states. States are happy to publicly endorse IHL treaties when it serves

36 W. BRADFORD, “In the Minds of Men: a Theory of Compliance with the Laws of War”, Arizona State Law

Journal 2004, 1257-1258. 37 E. GRIFFIN & B. ÇALI, “International Humanitarian Law”, in B. ÇALI (Ed.), International Law For

International Relations, 240-241.

24

their interests, but will defect from it just as swiftly to the same end.38 Realism refutes IHL is

based on humane considerations; they will only comply with IHL as long as this results in a

comparative advantage.39 In the context of an armed conflict the pressure of international

competition is so great, no factors of regime or policy would affect state behaviour. Compliance

will only be the outcome, if it would have been without the legal obligation as state security is

primary to calculations of state interest. Should any international institutions affect these

calculations, they do so because they mirror distribution of power.40

Second, liberalist approaches to IL believe in progress in international relations. International

cooperation coincides with national interests in certain issue areas. The liberal or democratic

regime does influence state behaviour, since it is associated with the rule of law. As such it

embraces the function of international legal institutions to facilitate said cooperation, overcome

problems of collective action and aggregate preferences according to which states comply.41

International law appears as a regime according to Krasner, which is a constellation of “(…)

principles, norms, rules and decision-making procedures around which actor expectations

converge in a given issue-area”. Non-compliance is contributed to the weakness of the regime.42

IL can also help clarify and enforce obligations.

Third, constructivist theory engages with IL on the premise that shared understandings of

proper conduct structure international relations. States internalise norms underlying obligations

because they proscribe what behaviour is appropriate and which is not.43 States evolve and

develop identities through interaction and the pursuit of collective interests. Constructivists

focus on socialization processes and the emergence and proliferation of international ‘norms’,

but tend to refrain from the “L-word” in their discourse.44 The English School is a subdivision

that combines the realist premise of power dynamics in international relations, but intend to

38 F. B. ADAMSON & C. L. SRIRAM, “Perspectives on international law in international relations”, in B. ÇALI

(Ed.), International Law For International Relations, 27-30. 39 E. GRIFFIN & B. ÇALI, “International Humanitarian Law”, in B. ÇALI (Ed.), International Law For

International Relations, 249-251. 40 J. D. MORROW, “When Do States Follow the Laws of War?”, American Political Science Review 2007, vol.

101., no. 3., 559-561. 41 F. B. ADAMSON & C. L. SRIRAM, “Perspectives on international law in international relations”, in B. ÇALI

(Ed.), International Law For International Relations, 27-30. 42 S. D. KRASNER, “Structural Causes and Regime Consequences: Regimes as intervening Variables”, in B.

A. SIMMONS et al. (Eds.), International Law and International Relations, Cambridge, Cambridge University

Press, 2006, 1-6. 43 J. D. MORROW, “When Do States Follow the Laws of War?”, American Political Science Review, 2007, vol.

101., no. 3., 559-561. 44 See supra 17, note 27.

25

include the societal dynamics of rule-governed behaviour.45 As IHL in particular concerns, it

appears as a normative framework for engaging in armed conflict. It provides vocabulary, rules

and persuasive dynamics to internalise the standards.46

Finally, the branch of rationalist institutionalism compromises and fuses several elements of IRT

together. Similar to realist theory the theory centres on states as rational actors in pursuit of

calculated self-interests. However, comparable to liberal tenets, institutionalism believes states’

pursuit of self-interest is reconcilable with international cooperation by creating institutions. As

these institutions and regimes grow and evolve they become more formal and institutionalised.

In contrast with constructivism, this process of institutionalization is not related to

internalization of standards for appropriate conduct. Instead its purpose pertains to facilitating

cooperation, optimizing processes through cost-reduction and providing interactive fora for

states to accomplish these goals.47 What does coincide with the constructivist views, is the

notion of shared understandings. These are required for actors to generate strategic

expectations of each other’s behaviour.

Rationalist institutionalist theories infuse international legal institutions with autonomous

explanatory power on the basis of this pursuit of interests.48 The rationalist strand focuses on

the most part on behavioural aspects of international relations or why actors portray behaviour

that is either in line with or deviates from international norms, rules and standards. As Keohane

posits “the most important functions of [international regimes] are to establish stable mutual

expectations about others’ patterns of behaviour (…)”.49 More concretely speaking to this

context, International Humanitarian Law is regarded as a political institution that serves to

embody guidelines for stable behaviour. These guidelines contain “a collection of considerations

beyond the control of an individual – norms, organizations, and formal processes and rules- that

impinge on his or her choice of action”.50 The incentives provided by the institution allow for

patterns of stable behaviour to emerge. All of the above reasons point to the potential of

45 F. B. ADAMSON & C. L. SRIRAM, “Perspectives on international law in international relations”, in B. ÇALI

(Ed.), International Law For International Relations, 37-41. 46 E. GRIFFIN, & B. ÇALI, “International Humanitarian Law”, in B. ÇALI (Ed.), International Law For

International Relations, 249. 47 F. B. ADAMSON & C. L. SRIRAM, “Perspectives on international law in international relations”, in B. ÇALI

(Ed.), International Law For International Relations, 34-36. 48 B. A. SIMMONS & R. H. STEINBERG, “Preface” in B. A. SIMMONS et al. (Eds.), International Law and

International Relations, Cambridge, Cambridge University Press, 2006, xxxii-xxxvi. 49 R. KEOHANE, “The Demand for International Regimes” in B. A. SIMMONS et al. (Eds.), International Law

and International Relations, Cambridge, Cambridge University Press, 2006, 23. 50 J. D. MORROW, Order within Anarchy, 5.

26

rational choice theory for this matter.51 The institutionalist paradigm lends itself quite nicely for

the application of Rational Choice Theory through game theoretic models. This methodological

pathway permits predicting behavioural decisions’ outcomes from which fruitful theoretical

conclusions can be drawn.

3.2 UNDERLYING ASSUMPTIONS

In the following pages I will develop a rational choice compliance model for IHL by extrapolating

a theoretical model for IL compliance in general. This will be largely based on Andrew T.

Guzman’s How International Law Works52 and James D. Morrow’s Order Within Anarchy53,

complemented with some insights provided for by other authors. A number of assumptions

underlie this model. As he duly points out, these assumptions are necessary for reasons of

feasibility, but leave room for complementary insights, gained from other points of view. In

particular IRT can offer additional contributions, as IL is only one factor influencing behaviour of

actors in the international community.54 Where it seems useful and feasible to incorporate some

of these other IR theorists’ insights, they will be put to use to complement the model as to

achieve a more general and complete model for IHL in the end. But first, let me introduce the

paradigm and state the basic assumptions.

1. The normative framework of IHL is operational: the threshold for an armed conflict has been

met

This dissertation seeks to develop a theoretical model for humanitarian law. This would only

make sense if those rules apply to the situations envisaged by the model. However, the

existence of an armed conflict is not always easy to establish in practice. The existence of

51 A variant on Rational Choice Theory posited by Keohane, nuances the rationality of actors’ behaviour by

emphasizing the constraining nature of the factors that govern the actors’ decision. Constraint-choice

analysis implies that environmental factors as well as other powerful actors have a determining impact on

policy choices made by the actors involved, if the actor considers the weight or importance of that factor

greater. This point of view seems useful in the realm of IHL compliance, since compliant behaviour is often

either induced or enforced by formalised sanction mechanisms or by pressure exerted in some form or other

from surrounding actors. - R. KEOHANE, “The Demand for International Regimes”, in B. A. SIMMONS et al.

(Eds.), International Law and International Relations, Cambridge, Cambridge University Press, 2006, 23. 52 A. T. GUZMAN, How International Law Works: A Rational Choice Theory, New York, Oxford University

Press, 2008, 16-70; 119-210. (Hereafter A. GUZMAN, How International Law Works). 53 J. D. MORROW, Order within Anarchy, 354 54 A. T. GUZMAN, How International Law Works, 216.

27

armed conflict is presupposed for any and all situations discussed. The establishment of an

armed conflict is less relevant here, as state participation is not hard to qualify. In the next

chapter I will elaborate further on the implications of the qualification of participants, the nature

of the conflict and the extent to which actors are bound by IL. For now, suffice it to assume the

threshold for armed conflict has been met and the state is under a legal obligation to uphold IHL

standards, whatever the extent, grounds or source of that obligation may be.

2. Human action is rule-governed. A fortiori state action is rule-governed IHL matters.

States are thought of as unitary actors. They constitute an aggregation of human actions.

Therefore state action is as much rule-governed as is human action. Norms and rules are

assumed to govern behaviour, because they categorise or order different options for actors

facing a choice.55 As they delineate the possibilities, they allow for actors, who are assumed to

be rational and self-interested, to apply rationality to their choice of action.56 The fact that IL

does matter remains the object of debate in legal scholarship and political science. Several

authors offer a wide variety of arguments testimony to this statement.57. For obvious reasons, it

would be of little use to develop a compliance paradigm for IHL under the assumption that it

does not matter.

3. Social processes can be best thought of in terms of (policy) choices made by states.58

55 F. V. KRATOCHWIL, Rules, Norms, and Decisions: On the Conditions of Practical and Legal Reasoning in

International Relations and Domestic Affairs, Cambridge, Cambridge University Press, 1989, 11. 56 ibid, 69-73. 57 Louis Henkin argues in his oft-cited work How Nations Behave that despite occasional deviations, IL

matters since, 1) despite a taxing political environment, compliance can be discerned; 2) mutual

interdependence creates state interests to comply with international obligations. – A. CHAYES, “Remarks by

Antonia Handler Chayes” in L. HENKIN How are nations behaving?, ASIL Proceedings 2002 206-207; Other

authors, f.i. Markus Burgstaller, argue on a consent-based theory that states are committed to compliance

since they choose to be bound. – M. BURGSTALLER, Theories of Compliance, 25; Thomas Franck posits the

reason why states abide by IL lies in the rules’ perceived legitimacy, derived from its determinacy, symbolic

validation, conceptual coherence and adherence to the organised normative hierarchy in the international

rule system. Whereas Abram and Antonia Chayes set forth the idea of an interactive process of justification,

discourse and persuasion awards rules with fairness that induces IL compliance. Harold Koh introduces the

concept of norm-internalisation through a transnational legal process, leading to compliance through

obedience. - H. KOH, “Why Do Nations Obey International Law?”, The Yale Law Journal 1997, 2628-2646;

Joel Trachtman argues that domestic political decisions determine whether states comply on the

international level. J. TRACHTMAN, “International law and Domestic Political Coalitions: The Grand Theory

of Compliance with International Law”, Chicago Journal of International Law 2010, 129. 58 A. H. KYDD, International Relations Theory: the Game-theoretic approach, 2015, Cambridge, Cambridge

University Press, 6.

28

Even an armed conflict is a manifestation of ‘social’ processes. Compliance decisions rely on the

choices states, soldiers and other actors involved make whenever a situation arises where a

conflict of interests exists.

4. States are assumed to be rational, self-interested actors able to identify and pursue their

interests, which are a function of their preferences.

The assumption of rationality goes beyond the notion of intelligent behaviour. The implication of

rational behaviour entails its motivation “by a conscious calculation of advantages that in turn is

based on an explicit and internally consistent value system”. Purported behaviour is thought of

to be rational when its aim is to maximise the valued decision outcomes according to the actors’

interests.59 Actors’ preferences co-determine how they value costs and benefits in the analysis

to reach a decision.

5. States have no innate preference for compliance with IHL.

State preferences are assumed to be exogenous and fixed; preferences are determined by their

dominant strategies: preference goes out to acquiring the greatest possible payoff. They are

unaffected by the rules’ legitimacy, past consent is not sufficient to ensure compliance and the

international norms are not assumed to be internalised. The fact that state preferences would

be fixed is obviously not true, but it would not be feasible to factor in all variables representing

public choice or domestic policy.

Under these assumptions the dynamics of IHL are conceived as stages in game theoretical

matrices.60 Possible outcomes are equilibriums, which consist of mutual best responses to each

other’s behaviour. In these games of strategy the interests and expectations of actors involved

determine what value they attach to possible outcomes of compliance or defection. In simple

cooperation games, the players share an interest in cooperation and defection always results in

mutually worse outcomes. Rational choice dictates cooperation will then be the outcome.

59 T. C. SCHELLING, The Strategy of Conflict, Cambridge, Harvard University Press, 1980, 4; 16. 60 For a comprehensible introduction to the basic notions of game theory, A Guide to Game Theory by Fiona

Carmichael offers a very adequate account. - F. CARMICHAEL, A guide to game theory, Harlow, Pearson

Education Ltd., 2005, 286.

29

Table 1: A simple coordination game61

In difficult coordination games, like the well-known Prisoner’s Dilemma, players can maximise

their total joint payoff through mutual cooperation, but each party does better for himself by

defecting. Consequentially, they would expect each other to pursue their own dominant

strategy. This would result in both parties defecting, which is a mutually worse, yet the most-

likely outcome in a one-shot game.

Table 2: A one-shot Prisoner's Dilemma

However, if a game is repeated over time, this has a significant impact on the players’

incentives: repetition creates the interest of future dealings, increasing interest in continued

compliance. This is often referred to in game theoretic terms as “the shadow of the future”,

taking into account considerations of a more long-term perspective.62 As international actors

deal with one another repeatedly over time during an armed conflict, this model is most

appropriate to apply to IHL.63 The model above shows what strategies players can pursue under

the assumption that they decide simultaneously on their actions. This might hold up for the first

stage of the repeated game that represents armed conflict: once the armed conflict “starts”

both sides have to make a decision to comply/defect at the outset. I will introduce an

alternative concept after clarifying three more assumptions first. An element characteristic of

Prisoner’s Dilemma is perfect information: actors observe each other’s behaviour completely.

This element is in no way characteristic for armed conflict, for reasons stated above. Hence this

aspect has to be addressed by another assumption:

61 Henceforth the payoffs of each actor will be visualised in these tables. The bold/italic font matches the

actors’ font in order to stand out better. Joint in the appendix the reader will find a detailed justification for

all the tables and models included in this paper. 62 F. B. ADAMSON & C. L. SRIRAM, “Perspectives on international law in international relations”, in B. ÇALI

(Ed.), International Law For International Relations, 34. 63 A. T. GUZMAN, How International Law Works, 25-32.

30

6. States are assumed to have imperfect information.

Actors have imperfect information regarding the other players’ payoffs, the content of legal

obligations and the other’s actions. The estimations of payoffs are usually based on a probability

distribution. As the game progresses, players can adjust their calculations from perceived

action. Legal standards are often the object of debate and interpretation, which causes some

difficulties for IHL in particular. Detailed agreements tend to relax this tension. Uncertainty

about actions can be the result of a lack of visibility or deception.64 These types of uncertainty

tend to reduce the incentive to comply. This leads us to another RCT assumption:

7. States cannot observe true willingness; they must estimate intentions on the basis of

observed behaviour.

At this point it is useful to stress the importance of shared expectations or common conjectures.

As Morrow rightfully points out: “What the actors think one another will do is as central to their

own calculations as their preferences over outcomes”.65 Actors will determine which equilibrium

they play based on what they believe is a credible response from the other. This observation

serves as well to explain the dynamic of deterrence: if a threat from one actor to decide on

certain behaviour is credible, the other will adjust his strategic expectations and choose a

different course of action.66 Another factor of influence is trustworthiness. Perceived in the

context of RCT trust can be defined as “a belief the other side prefers cooperation, to exploiting

one’s own cooperation” and mistrust is “the belief the other side prefers exploiting one’s

cooperation to returning it”.67 The payoffs of the Prisoner’s Dilemma compared to those in what

is known as the Assurance Game imply the compliance strategy equilibrium is dominant if

actors can trust on each other preference to comply.

Table 3: Assurance Games: trustworthy players

64 See supra 18, 2.3. 65 J. D. MORROW, Order within Anarchy: The Laws of War as an International Institution, New York,

Cambridge University Press, 2014, 20. 66 T. C. SCHELLING, The Strategy of Conflict, Cambridge, Harvard University Press, 1980, 6; 13-16. 67 A. H. KYDD, Trust and Mistrust in International Relations, Princeton, Princeton University Press, 2005, 6.

31

8. States have an interest in future dealings with their counterparts.

This interest is based on the actors’ concern to maintain good standing in the international

community. Guzman’s theory expands on three reasons why actors would comply with

international law in a repeated prisoner’s dilemma where the actors’ best strategies dictate

defection: reputation, retaliation and reciprocity –or what he refers to as the three R’s of

compliance.68 In order to influence each other’s behaviour, the impact of these three dynamics

on the different outcome payoffs must be big enough. As long as the total gain of compliance is

greater than the total gain of violation, rational choice dictates the actors will cooperate even if

the gain of violation is greater than the gain of compliance without these extra incentives.

Reputation is defined as “judgments about an actor’s past response to international legal

obligations used to predict future compliance with said obligations”. To extrapolate this

definition to IHL, we simply adduce the humanitarian nature of the obligations. These

judgments are estimations of actors’ true willingness to comply. A good reputation is deemed

valuable as it leads to easier, less costly cooperation in the future. Defection will harm an

actor’s reputation as it undermines the credibility of their commitments, thus inciting potential

defection from other players. The value awarded to reputation can differ greatly, depending on

several factors. One of these factors is the issue area. For IHL in particular reputation can only

do so much, because the stakes of survival are very high.69 However, in combination with other

incentives drawn from reciprocity or retaliation, the impacted payoffs might suffice to sway the

actor towards compliance. Moreover, by targeting specific aspects of a high-stakes

environment, the stakes can be compartmentalised, decreasing overall unwillingness to limit

one’s behaviour. For IHL, one example is provided by specific treaties on the use of chemical

weapons, which are generally seen as relative successes for IHL limitations.

The second “R” of compliance refers to retaliation. This figure implies a prior defection by one

player, to which the other replies with punitive action that is costly to the retaliating state. In

general international legal practice, such measures usually take the form of sanctions, in IHL

this usually means imposition of a greater force on the enemy. As these measures are costly,

such action is only rational if they generate another payoff for the retaliating state. That payoff

comes from another aspect of reputation, namely the reputation of making good on threats or

dealing harshly with violations. This will serve as a signal to future partners, increasing

credibility of future deterrence and expectations. If these measures can be removed once

mutual compliance is re-established, they can serve not only as a punishment but also as an

68 A. T. GUZMAN, How International Law Works, 33-69. 69 See supra, 18, 2.3.

32

incentive. In IHL retaliation used to be a traditional enforcement tool, yet by now developments

have sought to forgo this strategy.70

The third source of incentives is reciprocity. This implies reciprocal violation after prior defection

by an opponent, but these measures do not incur an additional cost on the reciprocating state.

They represent an adjustment in behaviour to the end of maximizing one’s payoffs in light of

new circumstances. This proves effective in bilateral contexts, especially when the gains of

single defection are small relative to gains from cooperation and violations are visible. This

dynamic is most promising for IHL as most armed conflicts represent a bilateral situation. Even

if more than two parties oppose one another, the number of sides can still be limited to two for

each compliance game in particular. Imagine intervention by a third party: if the party is neutral

to the cause of the conflict, chances are slim their objective will align with incentives to violate

IHL; if the intervening party is partial to the conflict, the natural dynamics of war will likely lead

parties to forge alliances or bandwagon, depending on which is perceived the bigger threat at

that moment.71 What is more difficult to conceive is the fact that the gain from single defection

would be small relative to the gains from cooperation. Here the relevance of asymmetry comes

into play. The gain of one-time defection will most likely depend on military power and prowess.

For the military inferior party one-time defection would be less likely to secure victory, which

would expose the party to an increased risk of retaliation or reciprocal noncompliance.

Therefore it would not be rational for the inferior state to choose to defect. For the military

superior party one-time defection could be very tempting, as this would significantly increase

chances of assured victory. Luckily this party can also benefit from this superiority for its

increased credibility of threats made to deter the other. This way he can a) avoid the

transaction costs of violation, and b) induce compliance from his counterpart through

deterrence. Should the threat lack credibility, reciprocity will fail to induce compliance.

Now it seems appropriate to introduce the concept of sequential moves.72 As the armed conflict

continues, and compliance decisions from opposing actors are interdependent the dynamic is

not represented by simultaneous decision. Actors respond to each other’s behaviour, which is

intrinsic for manifestations of behaviour to reciprocate or retaliate. The game theoretic model of

sequential moves allows for backward induction as a method to analyse which outcome is most

likely, taking common conjectures of what the other’s response might be into account.

70 J. D. MORROW, Order within Anarchy, 62. 71 S. M. WALT, The origins of alliance, 1978, Cornell University Press, 110-115. 72 F. CARMICHAEL, A guide to game theory, Harlow, Pearson Education Ltd., 2005, 79-103.

33

Model 1: Tree model for a sequential game

34

35

3.3 A RATIONAL CHOICE APPROACH TO IHL

OBLIGATIONS

The body of IHL norms mainly consists of two primary sources: treaties and Customary

International Law (CIL).73 Both will be discussed below from the perspective of RCT on

compliance. In the next chapter I will briefly touch upon soft law instruments regulating conduct

between states and NSA’s, but since the dynamic of soft law in intra-state relations is different

it will be omitted from this chapter. It would lead me too far to analyse this to its full extent

within the limited scope of this dissertation. 74

3.3.1 Conventional sources

3.3.1.1 WHY STATES WOULD FORMALISE AGREEMENT

The term agreement is used here to denote formal treaties and conventions. To enter into

agreements on humanitarian limitations on war can only be rational behaviour for states if this

would result in maximal payoffs, taking into account any transaction costs to achieve

cooperation. Once cooperation is achieved, securing its durability is of concern. Incentives to

violate can arise over time. 75 Why would states formalise cooperation, which requires a larger

transaction cost, if cooperation is mutually beneficial anyway? There is always the chance that

games change as they progress from a simple coordination game to a more complex game.

73 E. DAVID, Principes de Droit des Conflits Armés, Brussels, Bruylant, 2012, 54-71; For a list of treaties and

parties to them of the IHL body of instruments see the ICRC’s website: https://www.icrc.org/ihl; A study of

the ICRC on customary rules of IHL resulted in a list of 161 rules of CIL. See J. M. HENCKAERTS, “Study on

customary international humanitarian law: A contribution to the understanding and respect for the rule of

law in armed conflict”, IRRC 2005, 175-212. https://www.icrc.org/eng/assets/files/other/icrc_002_0860.pdf

, Annex p. 196-212. 74 There seems to be some progression towards informal sources of IHL in the works: see A. BUCHANAN &

R. KEOHANE, “Towards a Drone Accountability Regime”, Ethics and International Affairs 2015, 15-37; and

D. RICHEMONDE-BARAK, The 'Informal Law of War', working paper for the WBZ Berlin Colloquium on

Private Ordering and Public Authority, 2013, 25, Cited with the author’s permission.

https://www.wzb.eu/sites/default/files/u32/richemond-barak__the_informal_law_of_war_berlin.pdf; 75 This article goes on to categorise these hypotheses under three general hypotheses: 1) ambiguity and

indeterminacy of the obligation, 2) incapacity to perform and 3) contextual changes causing circumstances

to hinder performance. - A. CHAYES & A. HANDLER CHAYES, “On Compliance”, B. A. SIMMONS et al. (Eds.),

International Law and International Relations, Cambridge, Cambridge University Press, 2006, 69- 83.

36

Formalizing an agreement by convention requires a relatively small transaction cost and has the

potential to anticipate circumstantial changes that might influence the game. As such

formalization secures commitment. Treaties are also dynamic instruments that can be

amended, interpreted or exceptionally set aside if the need arises. A model for treaty constraint

can be represented in a two-stage game: in the first stage, states are unconstrained by legal

obligations. They negotiate with one another to reach agreement; in the next stage they have

to decide whether or not to comply with obligations previously agreed upon. As time goes by

and interactions repeat, the second stage is reiterated over and over. Moreover, the negotiating

process that precedes agreement will normally imply that long-term interests are promoted.

In the first stage states deploy their bargaining power to reach their greatest joint payoff in

comparison to their preferred outcome.76 Should there be no transaction costs to the obligation,

each would simply commit to the obligations resulting in the greatest payoff. However, each

obligation does have a corresponding cost. That is why states negotiate on the terms of

agreements. Guzman distinguishes two elements that are traded off to one another during

bargaining: substance, the set of formal promises that speak to actions states take in the

future; and form on the other hand, clauses that determine to which extent states pledge to

comply (possibility to make reservations f.i.) and enforcement clauses. What states’ preferred

outcome would be depends on how they value costs and benefits involved, which will influence

the agreement design. States observe each other in order to infer what merits they accord to

subject matters of design and form. More substance usually increases credibility but also

increases the cost, as flexibility diminishes. States would prefer for their own obligations to be

flexible, but the other’s to be rigid.

3.3.1.2 RISK AND RECIPROCITY

As stated, this model works under the assumption that state preferences are fixed. Guzman did

however realise this lack would cause a deficit for the model to adequately answer to reality. To

model these fixed preferences somehow, the concept of risk is used in RCT. The players’

attitude regarding equal payoffs for different outcomes is divided in three categories: risk-

aversion, risk-neutrality and risk-propensity. Risk-neutral players are indifferent between equal

outcomes when one is sure and the other unsure. Risk-aversion indicates that players prefer a

sure outcome to a gambled outcome if both have the same expected value.77 Most model states

as risk-aversive players; Guzman’s model thinks states to be risk-neutral. The amount of risk an

outcome bears can be managed in two ways. For one, by diversifying legal obligations and

76 T. C. SCHELLING, The Strategy of Conflict, 1980, Cambridge, Harvard University Press, 22-24. 77 F. CARMICHAEL, A guide to game theory, Harlow, Pearson Education Ltd., 2005, 123.

37

engaging several interests and projects simultaneously, states can achieve their goals through

different paths of cooperation. This way one objective does not stand or fall with a single

cooperation game. Second, states do not tend to make agreements on high-stakes matters -so

far there are no two states that promised each other never to go to war with one another

whatever the circumstances. In a high-stakes environment the risk is usually distributed by

cooperating on small areas of subject. Another factor relevant for the amount of risk is the

consequence of breach: enforcement or sanction increases the risk of violation, thus inducing

compliance. In high-stakes matters states do not tend to provide for strong enforcement

mechanisms, as this would imply them limiting their own margin for action. Moreover, the most

severe threat in the international order is outbreak of war, which has already happened in an

armed conflict. Hence, parties have no other means to enforce compliance apart from

reciprocating.

The dynamic of reciprocity is easy to conceive in a repeated Prisoner’s Dilemma game. The

assumption of perfect information however, does not hold up to model compliance in armed

conflict after. Problems of agency and noise disrupt the dynamic, causing reciprocity to be

irregular and disproportionate Reciprocity can be implemented on two levels to enforce

compliance: on the state level through policy and on the individual level through individual

actions. Individual reciprocation can lead to general reciprocation indirectly, should it trigger an

escalation spiral. Since actions of violation and reciprocity cover the same content, to

distinguish the one from the other is not an easy task. This purports to the need for shared

understandings to clarify thresholds and prevent misinterpretation of perceived behaviour. This

is exactly what norms can do: “shape action by changing the considerations in decisions.”78

3.3.1.3 COMMON CONJECTURES

For shared understandings and common conjectures to effectively predict behaviour, they must

meet a number of conditions.79 For one, the common conjectures must specify how states will

cooperate. Treaties can do so by putting in place thresholds of acceptable and unacceptable

behaviour. Unacceptable behaviour constitutes violation, which will incite reciprocal action. The

common conjecture must also specify how violations will be addressed. States must have

shared understandings of what reciprocal action is appropriate for violations. Finally, the

common conjectures must be mutual knowledge. To this end parties must signal their

commitment to upholding the agreed upon standards. In case of treaty law, ratification serves

as a strong public signal of intended commitment. To be effective signalled commitment must

78 J. D. MORROW, Order within Anarchy, 30. 79 ibid., 46.

38

be credible. This requirement creates incentives in turn to uphold the standards as the game is

repeated. If not, players will adjust their expectations accordingly, decreasing cooperation in the

long-term.

This dynamic explicates the potential of IHL to strengthen norms through specificity and public

commitment. Negotiation and ratification are useful tools in strengthening conjectures, yet they

are absolutely no guarantee for certainty. The problems of noise and imperfect information are

not solved through formal agreement. There will always be room for violation, be it intended or

otherwise. Moreover, treaty law can never exhaustively provide in prescriptions for every

possible contingency. This is where customary law enters the scene by providing “back-up”

principles to shape expectations. Naturally shared understandings as to the content and

operability of customary norms make lines less “bright”. The margin for different opinions on

what behaviour is appropriate and what constitutes violations is bigger. Nevertheless,

customary norms are still norms and exert a pull towards compliance as such. The following

paragraphs are devoted to explaining this dynamic.

3.3.2 Customary sources

The second primary source of IHL is CIL. The role of CIL consists of 3 functions that legitimise

its binding force: 1) it provides principles that undergird IL, including treaty law; 2) it plays a

supplementary role providing norms in case of lacuna; 3) doctrine has amply established that

CIL is a primary source of IL and is binding on all states, especially when the norms are

awarded the status of jus cogens, against which even the “persistent objector” doctrine

provides no recourse. From an RCT point of view the binding force of CIL is not generated from

explicit consent, but from a desire to protect reputation and answer to incentives from

reciprocity or retaliation threats. Traditional doctrine invests CIL with two constitutive

characteristics: opinio juris sive necessitatis and ‘general and consistent’ state practice. Guzman

points out that this definition is purely doctrinal in nature. The concept of opinio reflects the

conviction that standards and limitations are accepted from a sense of legal obligation. The

concept of state practice remains the object of debate, as there is no consensus about its

contours: how much, what kind and how widespread practice constitutes “consistent” practice.

Though a lot of IHL norms are considered to be CIL, amply providing for opinio juris, sufficient

state practice is often lacking. Yet this does not undermine the rule-like nature of customary

norms and the enhanced status attached to them.

This justifies the adaptation of the traditional definition to his functional definition of CIL.

According to his definition CIL consists of norms that, because they are considered to be law,

affect state payoffs more extensively. This results in a circular reasoning: if customary norms

39

affect incentives, they can be considered as customary. This constitutes a problem of

identification: to provide an objective criterion for identification, allows to test whether

behaviour is affected. This is especially hard for customary norms, which naturally grow over

time and depend on perception, which is invisible.

Let me continue with the reasoning that norms generate credible expectations of commitment

and will affect payoffs, thus affect behaviour. The amount of credibility depends on the level of

commitment. As treaties come with explicit consent and the will to cooperate, they are assumed

to generate a higher level of commitment than would obligations from other sources. Explicit

consent provides the counterparties of a stronger signal of commitment. If CIL is to be

distinguished from regular law by increased binding force, this too will generate a higher level of

commitment than regular norms would. Commitment in RCT can be regarded as a spectrum on

which norms of different natures take different positions. More commitment generates higher

expectations, affecting the cost of violation (be it derived from the objective payoff or

reputation, reciprocity or retaliation), thus affecting the payoffs. The magnitude and extent of

the consequences derived from binding force, is what separates CIL from other legal norms. The

obligations derived from each of these two sources generate a stronger pull towards

compliance, as the level of commitment in signalled more strongly. States are considered to be

more committed to upholding treaty obligations, than they are to upholding customary

obligations and more committed to customary obligations than to regular obligations.

Although the RCT dynamic is mostly reduced to bilateral relations, the dimension of general

consensus in the process of interaction in the international community comes into play. As more

actors share the belief that some rule is customary in nature, this will affect all of their

international relations by impacting their payoffs and behaviour. Subsequently when one state

estimates the value of their payoffs in a given compliance situation, their estimations will be

influenced by interdependence of other actors’ belief. The collective belief that a norm is

customary in nature will raise the expectation of commitment by the counterpart in question. In

short: if an entire region of states believe that and behave as if a legal obligation is customary,

other states will raise their expectations of any of them committing to said obligation. If a norm

is universally considered as customary, the expectations of commitment will be even higher,

thus the customary nature of the rule can be deduced from the extent to which it affects payoffs

universally. This provides an explanation for the dynamic of international discourse where

several actors try to influence which norms should be regarded as authoritative. As advocacy

and lobby groups try to convince actors that certain standards have customary status, the

actors they manage to convince will adapt their payoffs and behaviour accordingly.80 The

80 Think about the relation there could be between the practice of “naming and shaming” and reputational

effects f.i. - G. M. STEINBERG & A. HERZBERG, Best Practices for Human Rights and Humanitarian Fact-

Finding”, Leiden, Brill, 2012, 8.

40

doctrine of persistent objector can be explained from this framework too: if states manage to

convince others of not being bound by a certain rule, other will lower their expectations of

compliance and adjust their payoffs accordingly.

Finally, a remark on how opinio juris and state practice relate to one another. As perceptions

and convictions are immaterial, opinio cannot be observed as such. State practice however can

be observed by concrete actions or omissions that are forms in which these convictions

materialise. What then, constitutes trustworthy practice? As violations of customary norms do

still occur, inferring practice from them would immediately undermine the rule-like nature of the

violated norms. State declarations and statements cannot always be trusted to reflect actual

conviction as these concur with strategy and policy preferences. Alternatively, practice can be

deduced from the payoffs: if the payoffs dictate compliance, the binding force of a norm is

established; if violations occur despite the norms legal nature, this implies a cost for the

violating state (be it derived from reputation, reciprocity or retaliation). If there is no other

rational incentive, explanation or compensation for this cost, the customary nature of a rule can

be safely assumed under the rational choice paradigm.

The theoretical paradigm of RCT also fits for other theoretical approaches of CIL compliance,

such as the one offered by Verdier & Voeten81. They posit that the particular logic of CIL affects

interpretations of behaviour due to its universal nature, shared legal understandings and

institutional features. This paradigm leaves room for emphasis on the constitutive dynamic of

precedent, due to which “a state may comply because it knows its decision to defect creates a

precedent that may undermine a cooperative norm it values”. In addition they indicate that the

extent to which precedential concerns have an impact, depends on the features of particular

issue areas. One practical example illustrates this dynamic nicely: in 2011 the US government

contemplated cyber warfare strategies to disable Libya’s air defence system. Out of fear of

setting a precedent, they refrained from this venture.82 They too admit that the context of an

armed conflict raises difficulties for compliance as long-term consequences or reputational harm

will matter less. Short-term costs resulting from reciprocity and retaliation are more promising

to impact state behaviour in a high-stakes environment. However, the long-term consequences

cannot be set aside entirely, as enforcement mechanisms remain in place to constrain actors

legally and bureaucratically. It remains important not to forget these shared understandings are

81 P.-H. VERDIER & E. VOETEN, “Precedent, compliance and change in customary international law: an

explanatory theory”, AJIL 2014, 390 - 431. 82 P.-H. VERDIER & E. VOETEN, “Precedent, compliance and change in customary international law: an

explanatory theory”, AJIL 2014, p. 408; Original source: E. SCHMITT & T. SHANKER, “US Weighed Use of

Cyberattacks to Weaken Libya”, NY Times 2011 (Oct. 17), accessed online at:

http://www.nytimes.com/2011/10/18/world/africa/cyber-warfare-against-libya-was-debated-by-

us.html?_r=0.

41

more primitive in a sense. The concrete content is less well defined as treaty obligations. This

raises the risk of self-interested interpretations, causing behaviour to be perceived as violations,

even though the actors thought they were complying with their obligations.

3.3.3 Illustration of the dynamic

In order to best comprehend the pull of IHL legal obligations, be they treaty based or

customary, I will exemplify the dynamic by comparing the different possible constructions in

models. I will do so by drawing different pictures of how states would calculate their payoffs

under different conditions.83 Let me reiterate the variables the model seeks to account for:

- State preferences are modelled to be risk-aversive or risk-neutral;

- The model represents a repeated game;

- The model represents sequential moves;

- The actors are assumed to have imperfect information; noise and uncertainty hinder correct

estimations of behaviour;

- The payoffs represent the expected value of each outcome, taking into account transaction

costs and reputational or reciprocal costs;

- Depending on how actors estimate their counterparts’ payoffs, they will award an amount of

trust or mistrust on their expected behaviour

- The model can take into consideration elements of asymmetry, reflected in the payoffs.

The pull towards compliance that IHL generates essentially relies on the impact of what I will

refer to as “reciprocal payoffs”, which affect payoffs. The independent payoff is what value

states adhere to the possible outcomes without the impact of the “Three R’s” of compliance.

83 Please note that these conjectures might not answer to real-life examples in practice. Due to lack of

expertise and data on my behalf, the initial basic payoffs for military strategies cannot be generated that

simply. These conjectures serve the sole purpose of illustrating the dynamics of rational choice. I am aware

that this represents a weakness for drawing sound conclusions from this paper. However, any conclusions

drawn from it can still point to the potential value of further investigating these premises.

42

1. States are unconstrained by IHL obligations; there is no agreement.

Table 4: Payoffs for unconstrained states

The payoffs reflect that both states would be better off should they comply, instead of defecting.

However, each would do better than the other by defecting unilaterally. The common conjecture

that each other’s best outcome is unilateral defection, will lead both to defect. Since neither of

them is committed to cooperation, they have no reason to trust in the other’s mutual

compliance. To comply would leave one exposed to possible exploitation by the other.

Model 2: Sequential game for unconstrained states

The RCT paradigm dictates both states will defect due to the common conjecture that the other

side does better by defecting. The difference in outcome for mutual compliance and mutual

defection is rather small. The military inferior state would endure a slightly greater loss under

mutual defection, due to the greater power of the superior state.

2. States have an agreement on IHL obligations they have mutually committed to.

Table 5: Payoffs for constrained states

When both sides are constrained by legal obligations under IHL, their payoffs are affected by

the dependent payoffs. These reflect the gains and costs of reciprocating or reputational

sanctions. Each is assumed to gain 5 from compliance and -10 from defection. The gain is

43

smaller than the loss due to the transaction costs of implementing the defective strategy. This

would take each of them some effort in acquiring the necessary means and resources. The

payoffs still reflect that both states would reap higher outcomes under mutual compliance. The

difference with the outcomes for mutual defection is greater now. Defection has become more

costly.

Model 3: Sequential game for constrained states

The RCT paradigm would still dictate mutual defection, were it not that the game is repeated

and both sides have publicly signalled their intent to comply. As long as neither of them would

undermine the credibility of that promise (by defecting), they could maintain a stable

cooperative relation. The common conjecture resulting from the agreement will serve to

increasing the expectation of compliance.

3.3.4 Preliminary conclusions

It appears the commitment to legal obligations on the part of states could significantly impact

their payoffs and exert a pull towards compliance. Reciprocity and reputational sanctions

increase the cost of defection, making compliance look all the more appealing to settle for a

smaller individual outcome, while securing the maximal joint payoff. Ratification serves to the

end of supporting the credibility of said commitment.

Morrow’s research has led to great results through statistical analysis of treaty compliance by

testing data on violations, factoring in different variables such as the issue area of IHL, the

impact of noise and individual violations and regime type.84 The different views of realism,

liberalism and constructivism on the impact of IHL were tested against several hypotheses. The

results bear evidence that accord with and refute some stances the other three currents take.

84 J. D. MORROW, “When Do States Follow the Laws of War?”, American Political Science Review, 2007, vol.

101., no. 3., 559-572.

44

For one, it becomes apparent from the data analysis that unilateral compliance is rare. The

results bear evidence of treaty ratification affecting the level of compliance in at least three

manifest ways. First, his work indicates that joint ratification reinforces reciprocity, thus

increases compliance. As treaties can provide for both a signal of willingness to comply and

“bright lines”, reciprocation will address situations of violation more adequately. Second, as the

possibility of individual violation is greater, depending on the issue area, the level of compliance

tends to decrease. This indicates that agency and noise problems have a significant impact on

compliance. This aspect relates to the level of centralization of control over violations: an

individual soldier’s compliance decision relates more to treatment of civilians for instance, as it

does to the use of chemical weapons that is part of state policy. As for regime type, the results

indicate that democracies comply more when they are legally bound. However, in the

hypothesis of not being legally bound, democracies tend to comply less.

Evaluation of these results lead to conclude that the problems of agency and noise should be

addressed, which can probably best be done by formation and training. Evaluation in light of the

realist, liberalist and constructivist contentions, accords with the liberalist view on IL most. They

indicate that treaty ratification can in fact increase IHL compliance. This observation accords

with both the liberalist and constructivist position. The influence of regime type supports the

liberalist view too, where the constructivist point of view argues for norm-internalization. The

results do not support such hypothesis, as unilateral compliance is apparently rare. The dynamic

of reciprocity find ample support. This accords with the realist contention that reciprocity and

power affect interstate relations because IL is epiphenomenal to state power. The influence of

the institution finds support in the results as well, which contradicts realists’ rejection of the

potential of IL to constrain states behaviour.

3.4 ENFORCEMENT AND SUGGESTIONS FOR INCREASING

COMPLIANCE

3.4.1 Enforcement

One more aspect of IHL compliance merits a word in this context: implementation and

enforcement. Within the framework of IHL enforcement generally takes place on two levels. This

is a consequence of the actor-centric approach to IHL, regulating conduct for unitary actors,

states and groups of actors on the one hand and for individual actors on the other.85 First, as to

85 See supra 3.1.

45

violations of states, the mechanism of international state responsibility can entail a right to

reparation for the state violated against. Second, individual criminal responsibility for war

crimes and crimes against humanity seeks to penalise individual perpetrators through

adjudication.

Only limited success can be ascribed to these traditional mechanisms. A couple of obvious

reasons compel this statement. First, sanctioning violations presupposes their commission. It

seems that preventive compliance would yield far more valuable results. Second, the challenges

for IHL elaborated in the previous chapter produce uncertainty and a lack of visibility due to

which violations can remain undercover. Third, the mechanisms of state responsibility and

international adjudication are laced with well-known weaknesses. For international tribunals

their role is limited as there is no international coercive institution to enforce its decisions. As to

state responsibility, the normative framework needs strengthening, especially in cases where

states breach an obligation erga omnes. The argument here is an infringement of collective

interests of the community is not actionable under the state responsibility mechanism.86

Moreover the mechanism of state immunity undermines the effectiveness of state

responsibility.87 From a rational choice perspective it is difficult to imagine what payoff would

incite a state to uphold an obligation of reparation, once it decided to violate its legal

obligations. If the rule did not achieve to incite compliance in the first place, it is not likely to

produce an incentive to repair damages.88 Finally, this mechanism is only suitable for violations

that can be attributed to exercise of public authority. Violations by other than State-Actors fall

outside the scope of this mechanism. The aspect of enforcement of IHL obligations of AOG’s will

be discussed further in chapter four.

3.4.2 Possible areas of improvement

Suggestions in the literature to ameliorate IHL’s standing in the international legal order pertain

to several of these problems. One area of improvement deals with the substantive norms’ role

in RCT: clarification of norms and increasing transparency to ameliorate information would allow

for actors to make better decisions.89 Moreover the fact that legal rules tend to lag behind on

86 M. DUPUY, “The Deficiencies of the Law of State Responsibility Relating to Breaches of ‘Obligations Owed

to the International Community as a Whole’: Suggestions for Avoiding the Obsolescence of Aggravated

Responsibility” in A. CASSESE (Ed.), Realizing Utopia, 210-226. 87 P. GAETA, “Immunity of States and State Officials: A Major Stumbling Block to Judicial Scrutiny?”, in A.

CASSESE (Ed.), Realizing Utopia, 227-238. 88 A. T. GUZMAN, How International Law Works, 49-55. 89 A. T. GUZMAN, How International Law Works, 91-100.

46

societal evolutions poses an argument pro clarification of the body of IHL. 90 Monitoring and fact-

finding is an area that could remedy the lack of visibility of IHL violations. This mechanism could

be expanded to all multilateral treaties governing IHL limitations.91

Another suggested improvement pertains to formal enforcement through institutions. Cassesse

suggested a four-point plan to strengthen IL compliance through an international body

competent to determine breaches. To be successful this institution should provide for a follow-

up mechanism and recourse for individuals before an international court. Constitutionalizing a

principle of primacy of international rules constitutes the fourth element.92 Another suggestion is

to enforce international obligations before national judiciaries.93 Other concerns that have been

voiced purport to the need to strengthen the institutions already in place like the International

Criminal Court (ICC)94 and the International Court of Justice (ICJ)95 and to address the need for

coordination between international judiciary bodies.96 As to state responsibility, the limitations of

immunity should be reviewed and a mechanism for enforcing erga omnes obligations under

state responsibility could prove interesting for IHL. 97

90 N. MELZER, “Bolstering the Protection of Civilians in Armed Conflict”, in A. CASSESE (Ed.), Realizing

Utopia, 512-514. 91 Monitoring is a mechanism external to states and indicates pressure of the international community for

states to comply with international standards, in a way, which is acceptable to states that fear adjudication,

or inquiry might infringe on state sovereignty. It emanates from a collective interest; it is characterised by

state consent, automation, and continuity. Monitoring assessments do not have binding force. Cassesse

suggests these contentions are especially important for situations of armed conflict, as the provisions of art.

90 of the First Additional Protocol to the Geneva Conventions have never been put to use. Fact-finding is

characterised by legitimate installation by international institutions, its potency due to expertise, the

feedback to international institutions facilitating qualification in light of the international standards and

possibly follow-up measures. Suggestions to strengthen this mechanism relate to regulation, funding and

provision of experts at the ready. - A. CASSESSE, “Fostering Increased Conformity with International

Standards: Monitoring and Institutional Fact-Finding”, in A. CASSESE (Ed.), Realizing Utopia, 299-303. 92 A. CASSESSE, “Towards a Moderate Monism: Could International Rules Eventually Acquire the Force to

Invalidate Inconsistent National Laws?”, in A. CASSESE (Ed.), Realizing Utopia, 185-199. 93 Y. SHANY, “Should the Implementation of International Rules by Domestic Courts be Bolstered?”, in A.

CASSESE (Ed.), Realizing Utopia, 200-209. 94 A. CASSESSE, “The International Court of Justice: it is High Time to Restyle the Respected Old Lady”, in

A. CASSESE (Ed.), Realizing Utopia, 239-249. 95 W. SCHABAS, “The International Criminal Court: Struggling to Find its Way”, in A. CASSESE (Ed.),

Realizing Utopia, 250-260. 96 M. BENNOUNA, “How to Cope with the Proliferations of International Courts and Coordinate Their Action”,

in A. CASSESE (Ed.), Realizing Utopia, 287-296. 97 Since “the protection of human rights, the respect of cardinal principles of humanitarian law and respect

for the rights of the people – (are) rights the violation of which constitutes precisely a breach of obligations

‘owed to the international community as a whole’”, tackling this lack of actionability would indeed contribute

47

In 2011, the 31st International Conference of the Red Cross and the Red Crescent (ICRCRC)

initiated a consultation process including 145 states to work towards strengthening IHL

compliance.98 The result of this process was presented in a concluding report at the 32nd

ICRCRC in December 2015, where a resolution was adopted. The resolution adopted recalls the

guiding principles of the consultation process and recommends it be continued until the 33rd

Conference. They managed to establish general consensus on some key features. For one, the

need for a permanent forum where states can regularly discuss IHL has been amply established.

This institution would be given a number of additional competences, of which content remains

undecided. All documents stress this should be a legally non-binding mechanism, based on

dialogue, inclusion and cooperation. The process should be state-driven and consensus-based,

yet it should simultaneously avoid politicization. History has already taught us that these

characteristics provide rather shallow and weak institutions that tend to allow for politicization

to occur.99 The process has fairly little to show for after four years of deliberation: a resolution

with two articles, one on guiding principles and one to continue down the same path. Maybe it

should not surprise that such a process evolves rather slowly as it is a global intergovernmental

initiative on high-stakes matters. One might be doubtful of the efficiency of an informal, non-

legally binding mechanism in ensuring better compliance. Especially since the findings above

bear evidence that democracies comply better when they’re legally bound to. Nevertheless, this

initiative could most definitely function as a stepping-stone; it would not be the first time that

to inducing compliance with IHL. The author warrants caution in any such endeavour resulting in

progressive development of the law on state responsibility, since in the past recourse to the use of force to

the end of cessation of serious breaches of humanitarian law, has in turn led exactly to the use of force by

the international community. In the current state of international law no right to humanitarian intervention

exists and the legitimate use of force remains possible under strictly limited conditions. Moreover he warns

for three concerns that should be taken into consideration when reconsidering the ILC Articles on State

Responsibility: 1) this regime should only be effectuated in case of United Nations Security Council paralysis

in case the breach constitutes a threat to the peace; 2) third states should be able to invoke responsibility

and take countermeasures if the breach is not the object of any peaceful dispute settlement procedure; and,

3) any countermeasure should be proportionate to the gravity of the breach, not excluding the use of force,

but definitely limiting it to the stringent conditions of persistent wrongful conduct violating the most cardinal

principles of international humanitarian and human rights law. - ibid. 225-226. 98 All documents resulting from the consultation process can be consulted on the ICRC’s website at the

following page: https://www.icrc.org/eng/what-we-do/other-activities/development-ihl/strengthening-legal-

protection-compliance.htm; The documents resulting from the 32nd ICRCRC can be consulted here:

http://rcrcconference.org/international-conference/documents/. 99 Especially in the context of Human Rights protection, the institutional evolution towards universally

binding standards has shown the enforcement mechanisms in the several UN treaty systems proved weak

due to some of these characteristics.

48

an intergovernmental initiative grows out to become much more.100 From an institutionalist

point of view this could well be part of the process resulting in a formal institutionalised IO over

time.101

This concludes this chapter on state compliance with IHL from an RCT point of view. In the next

chapter this paradigm will be transposed to the level of NSA’s. The structure of chapter three

will be mimicked as best as possible. A first part is dedicated to demarcating the topic, since it

would be too ambitious an endeavour to exhaustively treat all categories of NSA’s. In the

second part I will revisit the RCT model by testing whether the assumptions hold up for AOG’s

as a primary phase. The next phase will evaluate whether IHL obligations of AOG’s to the extent

they are bound, generate comparable incentives that amount to an equivalent pull towards

compliance. Finally, the aspects of enforcement, challenges and suggestions for better

compliance will be discussed analogous to this chapter.

100 S. SMIS, C. JANSSENS, S. MIRGAUX & K. VAN LAETHEM, Handboek Mensenrechten: De internationale

bescherming van de rechten van de mens, Antwerp, Intersentia, 2002, 97-99; 354-357. 101 F. B. ADAMSON & C. L. SRIRAM, “Perspectives on international law in international relations”, in B. ÇALI

(Ed.), International Law For International Relations, 34.

49

4 Armed Opposition Groups and IHL

Compliance

4.1 EXTRAPOLATING THE PARADIGM TO NON-STATE

ACTORS

4.1.1 Delimiting the scope

Due to the limited scope of this dissertation, it will not be feasible to elaborate on all categories

of NSA’s and their incentives to comply with IHL. This dissertation will focus solely on AOG’s as

this seems to be the most stringent kind of actor to increase compliance from. Should an NSA

be more prone to violations of IHL, chances are they belong to the category of AOG’s. The

objectives of other categories of NSA’s are usually benevolent or at least neutral towards IHL

standards. I will justify this approach by explaining for the other categories of NSA’s why it

seems less urgent to deal with those. First, as regards to IO’s, they usually have plenty

incentive to abide by international law since they are governed by it themselves and their

competence emanates from the combined interests of their member states. Moreover, when

IO’s are endowed with international legal personality, this could imply their competence to enter

into treaty. Should the activities of an IO involve use of force, legal personality from a functional

point of view includes the possibility to enter into treaties governing said use of force. As

concerns reputational incentives, it seems unlikely that their impact on IO’s would be

insignificant, since they obviously have an interest in future dealings and perceived legitimacy.

Anything else would undermine their very existence. Third, it seems unlikely they would pursue

any goals or objectives that would imply the extension or intensifying violence during armed

conflict, on the contrary. Some of these arguments are valid for NGO’s too: they usually pursue

humanist goals in accordance with IHL,102 plus they have an interest in future dealings and

perceived legitimacy. As regards to Private Military and Security Contractors (PMSC’s), they

generally tend to endorse IHL obligations.103 Their objectives will be more commercial than

102 See X., International Non-Governmental Organisations Accountability Charter, 2005,

www.ingoaccountabilitycharter.org; See also G. M. STEINBERG & A. HERZBERG, Best Practices for Human

Rights and Humanitarian Fact-Finding”, Leiden, Brill, 2012, 1-2; 103 D. RICHEMOND-BARAK, “Applicability and Application of the Laws of War to Modern Conflict”, Florida

Journal of International Law 2012, 334; For a discussion on challenges for regulation of IHL arising from

50

humane in nature. However, they do have an interest in future dealings, which requires some

bureaucratic conditions such as certification be met.104 This could provide for an instrument to

influence their behaviour and create incentives to abide by IHL standards. If they are involved

in an armed conflict, this implies there is a contractual bond between themselves and a state. In

the context of an armed conflict, matters of compliance can probably best be approached

through conventional considerations, as enforcement would likely be a matter in the sphere of

domestic competences of the contracting state.

Should AOG’s have the same expected values as states for estimating their payoffs, the result

of any game would be the same as for inter-state conflict. Hence I will investigate where the

differences are likely to cause different incentives to comply or not. What is not the objective

here is to investigate which violations are more likely to be committed by AOG’s in comparison

to states. This dissertation wants to explore what differences lay in the incentive structures of

states vs. AOG’s and whether this leads to less IHL compliance in general. The behaviour in

different issue areas will fall outside the scope of this dissertation. Also, I will try to avoid

getting into any debate on the legality of certain actions, which is part of an entirely different

debate. It would lead me too far to account for the existence of violation or not, as the vague or

diffuse nature of possible criteria would obfuscate the aim at hand. The possible moves that

State-Actors (SA’s) or AOG’s can choose from are simply comply/defect.

4.1.2 The scope of application of IHL ratione materiae et

personae

The operability of IHL norms requires the actors to be legally bound by them in some way. The

extent to which NSA’s in general, and AOG’s in particular are bound by IHL is however, under

debate. In chapter three I have clarified the actor-centric approach to IHL, governing conduct

for both states and other group or individual actors participating in armed conflict.105 There are

several intertwined aspects that generate repercussions for the extent to which State and Non-

State Actors are obligated to uphold the laws of war. The scope of application of IHL requires

some clarification, since matters are fairly complicated due to the diverse interdependent

criteria. The next paragraphs are dedicated to elucidating the framework, whilst finding a useful

increased PMSC involvement see P. W. SINGER, Corporate Warriors: The Rise of the Privatised Military

Industry, New York, Cornell University Press, 2007, 348. 104 A. CLAPHAM, “The Regulations of Private Military and Security Contractors: Introductory Remarks”, ASIL

Proceedings 2013, 199-200. 105 See supra 44, 3.4.1.

51

subdivision for the overarching term “Armed Opposition Group”. The amalgam of goals,

institutional features, military capacity etc. of AOG’s vary to such extent that it would be

imprudent to make generalised contentions for the entire category all at once.

These criteria relate to the following aspects:

- The nature of the conflict and, its gravity and duration as regards NIAC’s;

- The qualification of the actors;

- The legal grounds for obligation;

- The source of the obligation;

- And, to some extent, the willingness of AOG’s to comply with IHL.

4.1.2.1 THE NATURE OF THE CONFLICT AND THE QUALIFICATION OF ACTORS

IAC’s

As concerns the nature of the conflict, different rules apply to IAC’s and NIAC’s.106 The existence

of an armed conflict is qualified as such on the basis of an objective standard for armed

confrontation, no matter the gravity or duration of the conflict. The threshold established in

Common Article 2 of the Geneva Conventions107, was purposely described rather vague to the

end of applying IHL to the biggest extent possible. The standard was intentionally set low and

the criterion relies on the empirical manifestation of hostilities.108 The concept of intra-state war

used to be the main object of IAC’s. States and IO’s are the main participating actors. They are

endowed with ILP and as such there is no problem with them exercising rights and duties under

IHL or enforcing them. In this case any operation or violation is either officially or de facto

imputable to the exercise of public authority.

Five other situations fall within the scope of an IAC: 1) the internal armed conflict, between a

state and an AOG is internationalised through third party intervention on behalf of a state, 2) or

an IO; 3) the internal armed conflict constitutes a liberation war between the state and the

106 Common Article 3 of the Geneva Conventions and Additional Protocol II are the main conventional

sources that apply in NIAC’s. As to the Rules of CIL, the study of the ICRC points out which rules apply in

NIAC’s. See supra 35 note 73. 107 Common Art. 2 Geneva Conventions: “In addition to the provisions which shall be implemented in

peacetime, the present Convention shall apply to all cases of declared war or of any other armed conflict

which may arise between two or more of the High Contracting Parties, even if the state of war is not

recognised by one of them. (…)”. 108 C. ZOLI, E. SCHNEIDER & C. SCHUSTER, “Armed conflict and Compliance in Muslim States, 1947-2014:

Does Conflict Look Different under International Humanitarian Law?”, North Carolina Journal of International

Law & Commercial Regulation 2015, 695-699.

52

national liberation movement109; 4) the internal conflict is a war of secession between

government in office and a group of people with control over some territory, which generates an

international nature of the conflict on the condition it succeeds; 5) the insurgency in a NIAC is

formally recognised to participate in the armed conflict. For any of these situations the entire

body of IHL instruments is applicable to the hostilities. This statement requires two more

nuances related to states’ conventional and customary obligations on the one hand and the

willingness of the AOG to uphold IHL.

NIAC’s

The category of NIAC’s appears to be a residual category. Any hostilities that qualify as an

armed conflict are governed by the IHL rules that apply in NIAC’s, and if the criterion of

prolonged duration is met, the provisions of AP II apply additionally. However, for the existence

of an NIAC, an additional criterion needs to be met concerning the gravity of hostilities. The

nature of an internal armed conflict requires confrontation of certain amplitude for the

application of IHL. The intensity of the conflict should be of comparable extent as would qualify

an IAC. In practice humanitarian intervention on behalf of the ICRC has been forwarded as

evidence to this requirement. The duration of an internal armed conflict determines what

provisions apply. Common Article 3 of the Geneva Conventions and Additional Protocol II (APII)

are the two main sources that apply in NIAC’s.110 They each have a different scope of

application: the AP II requires prolonged hostilities for its application, thus presents a more

narrow scope of application. The qualification of participating actors is more complicated in this

scenario. States and IO’s can also qualify as participants in NIAC’s. This generates no real

issues since they exercise some form of public authority. The same cannot be said for NSA’s.

For IHL to apply to insurgents, they should answer to the conditions set for constituting an

Armed Opposition Group. An AOG can assume rights and responsibilities under IHL if it meets

the twofold conditions of organisation and identifiable command. These conditions serve to

facilitate enforcement of violations of the IHL obligations.111

109 This category of AOG’s is assumed to fall under the category of belligerents for the purpose of this paper.

They can be bound by IHL if they fulfil the additional requirements of art. 1, § 4 First Additional Protocol. 110 Other provisions applicable in NIAC’s are: art. 19 Hague Convention for the Protection of Cultural

Property, 1954; art. 4 Statute of the International Criminal Tribunal for Rwanda, 1994; the Second Protocol

with the Convention Prohibiting Certain Conventional Weapons (CCW) of 1980, as amended in 1996; art. 8,

§ 2, c-f of the Stature of the ICC of 1998; art. 22 of the Second Hague Protocol for the Protection of Cultural

Property; the five CCW Protocols if the states have ratified art. 1 of the 2001 amendment to the CCW

Convention; art. 3 of the Statute for the Special Court of Sierra Leone, 2002; The rules of CIL that apply to

NIAC’s or all types of armed conflicts (See supra 35, note 73). 111 E. DAVID, Principes, 115-210.

53

There are however other criteria distinguishing types of AOG’s. Some categories have already

been mentioned: National Liberation Movements and a seceding group of people. This

distinction is based on the political objective of the insurgency. Another criterion for distinction

relates to the methods of warfare the insurgent group seeks to implement: guerrillas or

terrorists can be appointed as such. One more criterion has been brought forward in practice for

distinguishing AOG’s: recognition of formal participation in armed conflict. The consequences

are different for IAC’s and NIAC’s, therefore this criterion will be discussed separately.

4.1.2.2 THE CRITERION OF RECOGNITION

Formal participation in armed conflict generated consequences for determining which rights and

duties are bestowed onto the AOG’s. This concept of formal recognition used to provide

sufficient recourse for qualifying NSA’s participating in armed conflict. The AOG would

thenceforth be granted the fictitious personality of belligerents as an “ad hoc” state. The

outcome of the armed conflict would determine whether this status was retroactively invoked or

the NSA would actually become a new international legal person. Under the logic of IHL then

this categorization sufficed, since the act of recognition of an insurgency, which provided them

with the status of belligerents “internationalised” the armed conflict. Consequentially the entire

body of IHL instruments would apply to both the State- and Non-State actors involved.

However, there is no duty to recognise an AOG as such. States involved in the armed conflict

tend to withhold recognition, since this is perceived as admittance of weakness or undermining

of effective control.112 Recognition is however, not required to come from the opposing state

party as such. External recognition of AOG’s can crystallise the status of belligerency. Two

recent examples can be found in the cases of the Libyan National Transitional Council and the

Syrian Opposition Coalition.113 Hence this criterion proves useful for distinguishing categories of

AOG’s, especially because of the link with the nature of the conflict, which in turn determines

the extent of applicable obligations.

Paust distinguishes five types of AOG’s with formal status that have engaged in violence against

a state or another AOG.114 These types are not mutually exclusive. The fourth and fifth types are 112 I. DETTER, Law of War, Cambridge, Cambridge University Press, 2005, 42-43. 113 S. FOSTER HALABI, “Traditions of Belligerent Recognition: the Libyan Intervention in Historical and

Theoretical Context”, Am. U. Int’l L Rev. 2012, 372-389; S. TALMON, “Recognition of Opposition Groups as

the Legitimate Representative of a People”, Chinese JIL 2013, 220-226. 114 The first three types are: 1) ‘Nations’ which are “composed of a recognised group of people that may or

may not have a governmental process or a formal territorial base (…) [with] some form of recognition of its

status by other actors in the international community.” 2) ‘Peoples’ which are “[a recognised group of

people that may or may not have a governmental process of formal territorial base] that may or may not

54

especially relevant to this study: belligerents and insurgents, two technical statuses under IHL.

The status of belligerent requires five customary constitutive elements to be present: 1)

representation of an identifiable group of people, 2) the semblance of a government, 3) an

organised military force, 4) control over significant portions of territory and 5) outside

recognition as a belligerent, nation, or state.115 When these criteria are all met, the status of

combatant is awarded to the belligerent armed forces, in which case they enjoy protection

under IHL from anyone bound by it. In the context of an IAC, the parties involved are either

State-Actors or belligerents: the outside recognition as a belligerent or any higher status

explicates the international nature of the conflict.

An insurgency requires the same constitutive elements, with the difference that their status is

not recognised by an outside state actor. This category of AOG’s with some formal status

operates in NIAC’s, which can however be internationalised by third party intervention or

territorial extension beyond national borders. Insurgents cannot appeal to the status of

combatant; therefore they fall outside of the scope of these protective measures. The AOG’s

without formal participation in armed conflict are divided into five more types116, one of which

spikes the interest of this discourse: Terrorist Groups. They are defined by their choice to use

terrorism as a tactic of strategy, however a clear-cut definition remains absent in the realm of

IL. The application of IHL norms has proved tremendously difficult in the fight against

terrorism.117 The increased difficulty of applying the principle of distinction between combatants

share a number of common characteristics”. 3) ‘Tribes’ which are “an association of people (…) normally

more closely connected than a nation or people as such (…) usually smaller and [may or may not have a

territorial base]. – ibid., 274-278. 115 J. PAUST “Armed Opposition Groups”, in M. NOORTMANN et al. (Eds.), Non-State Actors in International

Law, Oxford, Hart, 2015, 279; original source: J. L. PAUST et al., International Criminal Law, Durham NC,

Carolina Academic Press, 200, 335. 116 Paust denotes 1) ‘Mercenaries’ as “persons specially recruited locally or abroad in order to fight in an

armed conflict (…), [which] takes direct part in the hostilities, is not a national or member of the armed

forces of a party to the conflict, has not been sent by a State which is not party to the conflict on official

duty (…) and is motivated (…) essentially by the desire for private gain” 2) ‘Pirates’ are “perpetrators [who]

engage in any illegal acts of violence or detention, or any act of depredation, committed for private ends…

directed on the high seas; (…) some (…) also have a purpose to oppose the power or the authority of a

state or other non-state actors.” 3) ‘Brigands, banditti and marauders’ are compared to pirates on land and

4) Organised Criminal Groups engage in (international) crime in an organised fashion and can occur as

actors in armed conflicts. – ibid., 286-290 117 Indeed, states have attempted to suspend their obligations under IHL in the fight against terrorism,

claiming that it does not apply. - F. B. ADAMSON & C. L. SRIRAM, “Perspectives on international law in

international relations”, in B. ÇALI (Ed.), International Law For International Relations, New York, Oxford

University Press, 2010, 28; D. RICHEMOND-BARAK, “Applicability and Application of the Laws of War to

Modern Conflict”, Florida Journal of International Law 2012, 329-340. This issue fits in neatly with related

55

and civilians has led the debate towards a notion of ‘unlawful enemy combatant’ when referring

to enemy targets of terrorist organizations.118 Moreover the application of the distinction

between IAC’s and NIAC’s, with implications for the applicable provisions, is challenged due to

the geographically widespread nature of terrorist attacks.

4.1.2.3 THE LEGAL GROUNDS FOR OBLIGATION

As to the grounds on which they are bound doctrine has offered that these AOG’s are bound by

IHL on different accounts. This explanation relies for the most part on the principle of legislative

jurisdiction, centring on the competence of the state to bind all individuals within its territory.119

From this premise ensues that the AOG’s that are bound by IHL under the several possible

hypotheses, are bound to the rules in force ratione loci in the territory where the conflict takes

place. In a complementary fashion four other bases explain why these actors are bound by IHL:

1) AOG’s that have expressly consented to be bound by the provisions of IHL, are bound due to

analogy with general treaty law120; The extent to which they are bound by IHL is then co-

determined by the obligations they committed too. 2) AOG’s that meet the criteria to represent

individuals in function of government are bound on the doctrine of succession of governments;

in that case they are bound to commitments of the previous government according to the

doctrine of state succession; 3) rules of IHL that are accepted to have the status of CIL are

binding on any and all AOG; and lastly, 4) rules of IHL reflecting general principles of IL are

binding on all AOG’s on the merits of that status.121

contentions regarding the abuse of humanitarian (or other liberalist) discourse by state actors for the sake

of protection of fundamental democratic principles to further the pursuit of national self-interest with very

perverse societal results to show for. These contentions accord with a realist-rationalist point of view as to

state foreign policy. However, from a constructivist point of view this can also set in motion a spiralling

dynamic of norm internalization as regards domestic policy. This illustrates how different takes on IRT and

IL have different contributions to offer and how social sciences are intertwined. For an interesting take on

how the Bush doctrine implemented liberal interventionism in the “War on Terror” and the repercussions

thereof in the US and the Middle East see D. KUMAR, Islamophobia and the politics of Empire, Chicago,

Haymarket Books, 2012, 123-136. 118 D. RICHEMOND-BARAK, “Applicability and Application of the Laws of War to Modern Conflict”, Florida

Journal of International Law 2012, 341-354. 119 S. SIVAKUMARAN, “Binding Armed Opposition Groups”, International and Comparative Law Quarterly

2006, Cambridge, p. 381. 120 As is comprised in articles 34-36 of the Vienna Convention on the Law of Treaties. 121 S. SIVAKUMARAN, “Binding Armed Opposition Groups”, International and Comparative Law Quarterly

2006, Cambridge, p. 371-381.

56

4.1.2.4 THE SOURCE OF THE OBLIGATION

From the previous section it became clear that AOG’s could be bound by IHL on the merits of its

customary status. As for conventional sources, the way ILP traditionally is conceived prevents

NSA’s from acceding to formal international agreements.122 Concerns have been voiced in

scholarly literature about the wedge between “law-takers and law-makers”. There is a growing

discourse in favour of adjusting the concept of International Legal Personality, in such a way

that NSA’s could play a greater role in obtaining international rights and duties, especially in the

context of IHL obligations.123 The notion of explicit consent to legal standards appears to have

an impact on compliance behaviour. Initiatives to allow for AOG’s to voluntarily commit to IHL

standards have had some success. The features of organizational capacity and centralization are

apparently crucial for success in this area.124 Voluntary consent allows for AOG’s to take

ownership of their obligations, increasing their incentive to comply.125 Informal arrangements

can obviously pose no problem for NSA’s, seeing that this requires no specific traits or

conditions. In light of both trends, an emerging tendency can be distinguished in the

development of IHL soft law.126

As for the extent to which states are bound to IHL, their obligations are co-determined by which

conventional sources they ratified and which customary sources they have not persistently

objected to. This nuance puts into perspective the earlier statement that states are bound to the

full extent of IHL in IAC’s. The same goes for their obligations in NIAC’s, but the body of IHL

rules that might apply is then limited to the rules indicated above.

122 G. D. SOLIS, The Law of Armed Conflict: International Humanitarian Law in War, Cambridge, Cambridge

University Press, 2010, p. 157. 123 J. E. NIJMAN, “Non-State Actors and the International Rule of Law: Revisiting the ‘Realist Theory’ of

International Legal Personality in M. NOORTMANN & C. RYNGAERT (Eds.), Non-State Actor Dynamics in

International Law: From Law-Takers to Law-Makers, Farnham, Ashgate, 2010, 109-119; M. NOORTMANN,

“Understanding Non-State Actors in the Contemporary World Society: Transcending the International,

Mainstreaming the Transnational of Bringing the Participants Back In?”, in ibid.; 160-166; J. D’ASPREMONT,

“International Law-Making by Non-State Actors: Changing the Model or Putting the Phenomenon into

Perspective”, in ibid., 180-182. 124 M. KORNPROBST, “Non-State Actors in International Relations” in M. NOORTMANN et al. (Eds.), Non-

State Actors in International Law, Oxford, Hart, 2015, 300. 125 E. DECREY WARNER, “Engaging with Non-State Armed Groups: Lessons Learned”, 36th Round Table on

Current Issues of International Humanitarian Law, “Respecting IHL: Challenges and Perspectives”, Institute

of International Humanitarian Law San Remo, 5-7 September 2013, 1. 126 See supra 35, note 74.

57

4.1.2.5 THE WILLINGNESS OF AOG’S TO COMPLY WITH IHL

For the extent to which states are bound by IHL the dimension of reciprocity arises here in a

different light. Under the Rational Choice paradigm, reciprocity appears as a characteristic or

corollary of IHL obligations: compliance behaviour can be induced by (threatening with)

reciprocal defection. This possibility needs to be nuanced in light of the Geneva Conventions.

Common Article 2 of the Geneva Conventions stipulates under which conditions States bound by

the Conventions have legitimate recourse to reciprocal defection.127 The conditions for reciprocal

defection limit this possibility to the context of an IAC, in which State-Actors are confronted with

an Armed Non-State Actor that neither accepts, nor applies the law.128 Even under these

circumstances however, the state is bound to some cardinal principles. These cardinal principles

pertain to the growing consensus on a Minimum Standard of Protection. The doctrinal concept of

jus cogens explicates the vital importance of cardinal principles. David distinguishes three

arguments pro this status based on the nature of IHL, the fact that it meets the criteria for jus

cogens,129 and elements of intangibility of certain IHL obligations.130 This implies that in an IAC

the state is bound to the full extent of its humanitarian obligations, if the AOG accepts and

applies them reciprocally. Furthermore, this implies that in a NIAC it is irrelevant whether or not

the AOG abides by IHL for determining the extent to which the State is bound. Whether this has

an impact on compliance behaviour according to RCT, I will investigate further ahead.

This analysis might have sidetracked the reader from the compliance ‘train of thought’.

Nevertheless this was a necessary evil in order to accurately specify which AOG’s compliance

behaviour will be scrutinised.131 Allow me to recapitulate briefly: if AOG’s are bound by IHL, they

are bound to it to the same extent as their state counterparts. There are in general three

constellations that result in IHL application. The first category concerns intra-state IAC’s. The

compliance behaviour for this type of conflict has been treated in chapter three. The second

127 See supra 51, note 107. 128 D. RICHEMOND-BARAK, “Applicability and Application of the Laws of War to Modern Conflict”, Florida

Journal of International Law 2012, 330-341. 129 Art. 53 VCLT: “A treaty is void if, at the time of its conclusion, it conflicts with a peremptory norm of

general international law. For the purposes of the present Convention, a peremptory norm of general

international law is a norm accepted and recognised by the international community of States as a whole as

a norm from which no derogation is permitted and which can be modified only by a subsequent norm of

general international law having the same character.” 130 These three elements pertain to expressions seemingly aimed at reinforcing the imperative nature of the

obligations, the fact that reprisals for IHL violations are forbidden and the fact that violations qualify as

criminal offences. - E. DAVID, Principes, 107-115. 131 The scope of application of IHL might not be very transparent to all readers, since there are many

variables determining the extent to which parties are bound. For a simplified schematic overview on this

matter, please consult the appendix.

58

category concerns NIAC’s pur sang. If the AOG meets the requirements (sufficient organisation

and an identifiable authority), and the conflict meets the criteria of gravity (and duration), the

IHL obligations for NIAC’s are in force. This category will be referred to as insurgents. The third

category concerns internationalised NIAC’s. In this case the AOG is recognised to formally

participate in the armed conflict, due to which the full body of IHL standards applies. That is,

unless the AOG resolutely rejects the norms of IHL, in that case the state can reciprocally

suspend most of its own obligations. This category will be referred to as belligerents. The

theoretical paradigm of RCT compliance will now be analogously applied to this second and third

category of AOG’s in the following pages.

4.1.3 Do the assumptions hold up?

1. The normative framework of IHL is operational: the threshold for an armed conflict has been

met.

This assumption bears no further explanation regarding the threshold of armed conflict. It

remains a given that the actors find themselves in such a situation. The fact that IHL applies

and to which extent the actors are bound is clarified in the preceding paragraphs.

2. Human action is rule-governed. A fortiori AOG action is rule-governed: IHL matters.

This assumption can be transposed self-evidently and requires no further elaboration.

3. Social processes can be best thought of in terms of the choices made by rational AOG’s

In this context all actors involved will be assumed to pursue their goals rationally. One might

argue that AOG’s proclaiming violence are irrational, because their message can be deemed

incomprehensible or unreasonable from an ideational or metaphysical point of view. The

strategy of civilian victimisation for example, a strategy to intentionally direct violence against

the population is “a short term strategy (…) rational to the extent that it helps insurgents stave

off collapse and denies victory to the government.”132 The Lord’s Resistance Army led by Joseph

Kony went on a civilian killing spree in 2008 when the Ugandan government implemented a

fierce organised military strike. Yet the rationality requirement does not depend on the content

of reasoning. What is common is strategic behaviour: the wish to make decisions so that they 132 R. M. WOOD, “Understanding strategic motives for violence against civilians during civil conflict”, in H.

KRIEGER (Ed.), Inducing Compliance with International Humanitarian Law: Lessons from the African Great

Lakes Region, Cambridge, Cambridge University Press, 2015, 15; 31-32.

59

result in the best possible outcome to the ends envisaged. This commonality is what justifies

awarding rationality133 to all actors involved in armed conflict, regardless of their proper

objectives. Their views will only affect incentives and how they value the outcomes and payoffs,

resulting in different equilibriums. An outcome is only irrational under RCT when external

factors disrupt the decision-making process (imperfect information, faulty calculation of

expectations or intentions, noise…), which cancels out any accusation of irrationality at the

address of the actor himself.134 The chosen strategy was rational, but the eventual outcome less

than ideal in such a scenario. The self-interest of states is an interest abstracted from its

collective interests as a nation. Likewise, self-interest of AOG’s is not assumed to be an egoistic,

individual interest. It might as well be identified with a common goal or policy.

4. AOG’s are assumed to be rational, self-interested actors able to identify and pursue their

interests, which are a function of their preferences.

5. AOG’s have no innate preference for compliance with IL.

This assumption can also be transposed: AOG’s are assumed to prefer the strategy leading to

the greatest possible payoff. This assumption is no more in accordance with reality as it is for

states, but the line of reasoning of the RCT paradigm requires objectifying this element just the

same. In reality AOG’s preferences will be affected by factors such as but not limited to policy,

ideology, history etc. For clarity’s sake, the inverse assumption is equally fundamental for the

model: AOG’s have no innate preference for violation of IHL. Rational choice dictates the actors

choose in accordance with the equilibrium strategy for the greatest possible payoff.

In the previous chapter state preferences were modelled by the concept of risk-aversion or risk-

neutrality. What preference AOG’s have as regards risk is not evident to establish. It would

seem imprudent to assume they are risk-aversive or risk-neutral since I have no ground to

found this statement on. In my research I have neither incurred evidence that AOG’s would be

risk loving, but the exclusion of the possibility feels wrong intuitively speaking. Especially since

our society has regular encounters with very rogue seeming AOG’s. I do realise this is not a

scientifically qualified contention. Therefore the aspect of attitudes towards risk will be omitted

from any models. It will suffice to bear in mind the underlying assumption of fixed preferences,

dictating how payoffs are valued.

133 Rationality in cognitive psychology is defined as internally consistent and logically coherent, and not

whether a person’s beliefs and preferences are reasonable or in tune with reality. Although one can make an

effort to make very calculated decisions, people are never 100% rational. – D. KAHNEMAN, Thinking Fast

and Slow, London, Penguin Books, 2011, 269; 411. 134 T. C. SCHELLING, The Strategy of Conflict, 1980, Cambridge, Harvard University Press, 15-18.

60

6. AOG’s are assumed to have imperfect information.

7. AOG’s cannot observe true willingness; they must estimate intentions on the basis of

observed behaviour.

These two assumptions are consequential to a battlefield dynamic and any party involved in

armed conflict is subject to it as such. Extrapolating these assumptions to AOG’s has no

implications for the theoretical model.

8. AOG’s have an interest in future dealings with their counterparts.

This last assumption is essential for the model of RCT developed, since it is this aspect that

impacts decision behaviour most obviously. The costs associated with the three R’s affect

payoffs and incentives as such. It seems this will evidently hold up for any AOG. Depending on

whether or not their endgame is to participate in the international order as it is construed, they

should award different values to reputational costs. On the other hand, just as states could

value a different type of reputation than the reputation to comply –dealing harshly with

violation f.i. One could posit that AOG’s that do not value a reputation to comply, inversely

value a reputation not to comply out of dissidence or rejection of the principles. Under this

hypothesis they would value an outcome following defection with no or low reputational costs

and vice versa, estimate an outcome following compliance with a high reputational cost. As the

theoretical paradigm hinges on this assumption, the theoretical model for compliance in an

armed conflict involving state and AOG actors needs to account for this.

To extrapolate these findings into game theoretical models would yield four sets of models, two

for each kind of armed conflict. In an IAC one model would project payoffs and strategic

decisions for states on the one hand and belligerents that accept and apply IHL. Another model

would project behaviour for states and AOG’s that reject IHL. Here the different extent of state

responsibilities will also affect outcomes. Both situations would translate to two models for

NIAC’s as well, after all the intentions of AOG’s in NIAC’s might not be relevant to the extent of

state obligations, but they surely would affect the AOG’s incentives to comply or not.

The problem here is that the assumption that AOG’s have an interest in future dealings will

probably not hold up for AOG’s that resolutely reject IHL entirely135. Imagine a seceding

135 According to Griffin and Çali possible reasons for lack of incentive to comply may be the lack of interest

in international legitimacy or recognition, collapse of mutual trust, escalation of violence, lack of checks and

balances within a non-state armed group, the inability of the AOG to discipline its members, the lack of

international pressure and monitoring or the existence of a very unequal power relationship between parties

61

opposition party attempting to achieve state recognition in the international order. It would

seem they have high stakes in abiding by international law, because of the general principle

stating that states have a duty not to recognise any situation that resulted from breach of IL.136

Imagine on the other hand an AOG with a program based on an ideology that fundamentally

rejects the international order as is and implements military strategies based on terror, crimes

against humanity or other atrocious acts of violence. The AOG ISIL, for example, has blatantly

flouted many standards of IHL in committing some of the most atrocious war crimes such as

genocide.137 It would seem an incentive to uphold IHL standards is hard to configure on such

program. Should they want to enhance and legitimise their position in the international order,

rejecting IHL would probably not be a strategically justified move. The theoretical models for

AOG’s that reject IHL would therefore not generate plausible outcomes and must be set-aside

entirely for both IAC’s and NIAC’s. This leaves us with two models left: one for IAC’s between

states and belligerents, and one for NIAC’s between states and insurgents, both AOG’s assumed

to have an interest in future dealings, since they choose to accept and apply IHL standards even

though their obligation to is under debate. In the next paragraphs these models will be tested to

see what impact IHL has on the compliance behaviour of belligerents and insurgents.

4.2 APPLYING THE RATIONAL CHOICE MODEL FOR IHL

COMPLIANCE TO AOG’S

The previous paragraphs have served to establish that the assumptions under the rational

choice model persist for insurgents and belligerents. In what follows the dynamic established in

chapter three will be transposed to belligerents and insurgents. Doing so I will try to recreate

the methodical structure as well as possible. The eventual goal is to theoretically estimate how

payoffs for belligerents and insurgents are affected by the pull towards compliance that

humanitarian legal obligations exert.

to the conflict. - E. GRIFFIN & B. ÇALI, “International Humanitarian Law”, in B. ÇALI (Ed.), International

Law For International Relations, 253. 136 Art. 41, (2) ILC articles on State Responsibility; See also S. TALMON, “The Duty not to “recognize as

lawful” a Situation created by the Illegal Use of Force or other Serious Breaches of a Jus Cogens Obligation:

an Obligation without Real Substance?” in C. TOMUSCHAT & J.-M. THOUVENIN (Eds.) The Fundamental

Rules of the International Legal Order: Jus cogens and Obligations Erga Omnes, Leiden, Martinus Nijhoff,

99. 137 N. A. ENGLEHART, “Non-state Armed Groups as a Threat to Global Security: What Threat, Whose

Security?, Journal of Global Security Studies 2016, 1.

62

4.2.1 Conventional sources

Belligerents and insurgents are bound by the treaties in force within the territorial boundaries of

the conflict, based on the legislative jurisdiction of the territorial state to bind all subjects within

its jurisdiction. Insurgents are bound to a lot less conventional obligations as recorded in the

previous section. Agreements can serve the same purpose of clarifying the framework for

cooperation between State-Actors and AOG’s. States that need to negotiate and enter into

treaties are confronted with a transaction cost of achieving these actions. Since AOG’s do not

participate in these dynamics, they incur no such transaction cost. This implies they have even

more to gain by cooperating. Earlier I pointed to the new trend of informal agreements between

State and AOG’s as to achieving better compliance. These soft law instruments will be briefly

discussed, since they are the actual equivalent of treaties for states: AOG’s are required to

actually negotiate and consent to these instruments. This will in turn imply a transaction cost,

but in any case it will probably be smaller than the transaction cost for states ratifying

multilateral treaties. These informal agreements are thought to exert a greater pull towards

compliance, since the AOG’s can take ownership of their commitments.138

This pull towards compliance will be enforced by the dynamic of reciprocation in the state-AOG

relation as well, since the same circumstances of armed conflict are present.139 Neither can

threaten the other with anything worse than going to war. The dynamic of reciprocity is

susceptible to noise, uncertainty, the problem of agency and imperfect information. This will

probably have a greater impact in the state-AOG relation, due to apparent asymmetry. AOG’s

normally do not reside over the same kind of resources as states do. The better the AOG is

organised, the less these issues will cause problems on the battlefield. If the AOG’s forces are

well trained, subject to discipline, well informed and aware of the policy implications of

upholding the IHL obligations, then fewer violations will either occur, or be mistakenly perceived

as such. If not, the risk of inadvertent or individual violations might trigger violence to spiral out

of control. Hence the notions of shared understandings and common conjectures are also

necessary in this relation. Treaties and agreements can clarify the content of the obligations, in

order to strengthen the shared understandings.

The common conjectures are subject to the same conditions for effectively estimating each

other’s behaviour. Since the same treaties are in place as would be otherwise, they will provide

the same threshold for appropriate behaviour and appropriate reciprocation. Any other

agreement can further clarify if the parties want to. What is different for AOG’s is that just as

138 See supra 35, note 74. 139 E. GRIFFIN, & B. ÇALI, “International Humanitarian Law” in B. ÇALI (Ed.), International Law For

International Relations, 253.

63

they do not participate in negotiation, they do not publicly ratify treaties as official members.

For the dynamic of reciprocity to effectively work, AOG’s should provide their state opponent

with an equivalent public signal of their willingness to comply. If the common conjecture to

comply is not mutual knowledge, this would create incentives to unilaterally defect. The AOG’s

promise to comply must be credible for compliance to be mutual. As the conflict progresses and

the game is repeated, compliant behaviour serves to provide the common conjecture with

credibility.

Nevertheless the dynamic is still subject to uncertainty and noise. Efforts to limit this to the

smallest extent possible can contribute to increasing compliance. The more precise the content

and the shared understandings, the less violations will occur or lead to reciprocal violations.

However, the instruments cannot foresee all eventualities. Again CIL provides a supplementary

mechanism for actors to fall back on when no agreed upon standard is set.

4.2.2 Customary sources

AOG’s can effectively find recourse to rights and obligations under customary law. CIL will too

exert a pull towards compliance, but it remains smaller than with conventional obligations.

Since the behaviour of AOG’s does not immediately present state practice, they are not as

closely involved in the development of CIL. Should the AOG be successful in the outcome of the

armed conflict, it might do so in a later stage. Their decisions at the time of the conflict can

have a postponed effect on their “state practice to be”. At the time of the armed conflict it

would be more likely the AOG’s decisions are influenced by the “advocacy” side of the CIL

dynamic. Their compliant behaviour could be constitutive for convincing the international

community of the authority a certain rule should be awarded. Since AOG’s role in the CIL

process is rather limited, it will suffice to leave it at this short observation. In the following

pages I will attempt to illustrate the dynamic of IHL obligations of AOG’s in the same game

theoretical models of chapter three.

64

4.2.3 Pouring the info into illustrative models

I will again reiterate the modalities the model seeks to account for before I proceed onto

conjectures of AOG’s payoffs.

- AOG preferences are assumed to be exogenous and fixed, but no assumption on the attitude

of the AOG towards risk will be applied;

- The model represents a repeated game;

- The model represents sequential moves;

- The actors are assumed to have imperfect information; noise and uncertainty hinder correct

estimations of behaviour;

- The payoffs represent the expected value of each outcome, taking into account transaction

costs and reputational or reciprocal costs;

- The model reflects asymmetry in the relations between states and AOG’s.

In order to have a basis for comparison, the first set of payoffs represents the objective payoffs

the AOG would estimate, would it be unconstrained by IHL. The model here should be different

for belligerents and insurgents. The analysis of the application scope pointed out that states can

reciprocally defect if the belligerent does in an IAC. Thus we depart from two sets of payoffs for

each type of AOG.

1. The state is constrained by its legal obligations in a NIAC; the insurgent is not.

Table 6: Payoffs for unconstrained insurgent

The payoffs for the constrained state are based on the payoffs in Table 5: Payoffs for

constrained states. The constraint is reflected in the dependent payoffs, which are not

represented for the insurgent. The payoffs for the insurgent reflect asymmetry in the payoffs for

unilateral defection and mutual defection. Both parties are better off under mutual compliance,

but best of under unilateral defection. The insurgent estimates mutual defection with a higher

outcome than unilateral compliance on his part.

65

Model 4: Sequential game for unconstrained insurgent

The total payoffs reflected in the tree model indicate that mutual compliance would be the

strategy resulting in the maximal joint payoff. However, as both are unconstrained, neither can

trust the other to comply. There is no agreement on cooperation, so they expect one another to

defect, since unilateral defection is the dominant strategy. Therefore, mutual defection will be

the outcome.

2. Both the state and the insurgent in the NIAC are bound by legal obligations under IHL.

Table 7: Payoffs for constrained insurgent

The total payoffs reflect that defection becomes more costly. The state incurs a cost of -10 for

defecting. For the insurgent defection requires a higher transaction cost, thus the cost is

estimated at -15, since they usually have less military resources.

66

Model 5: Sequential game for constrained insurgent

In the sequential model with the total payoffs, the implications of obligation for mutual

defection are quite clear. However, as long as both sides expect each other to pursue their own

dominant strategy of unilateral defection, mutual defection remains the outcome. In the

equivalent Model 3: Sequential game for constrained states, the common conjecture is shifted

to credible commitment to compliance, because both states have publicly ratified the

agreement. Insurgents do not automatically convey such a signal of commitment. For durable

cooperation to be established, the insurgent will have to convey its intent to comply though

some means of communication. In that case, the equilibrium of mutual compliance can be

sustained

3. Neither the state, nor the belligerent is constrained in an IAC.

Table 8: Payoffs for unconstrained belligerent

In this table the payoffs for the state are based on Table 4: Payoffs for unconstrained states. In

an IAC the state can suspend its obligations if its non-state counterpart rejects IHL. The payoffs

for the belligerent are based on Table 6: Payoffs for unconstrained insurgent.

67

Model 6: Sequential game for unconstrained belligerent

Under this constellation the same dynamic works as under Model 2: Sequential game for

unconstrained states. The main difference is that the belligerents’ payoffs represent

asymmetrical payoffs better. This does not have an immediate impact on the outcome strategy

though. The outcome will be mutual defection, since neither can rely on the common conjecture

that the other will comply. They will prefer to settle for the smaller payoff, this way preventing

the other from obtaining a higher payoff and simultaneously safeguard against the risk of being

exploited through unilateral defection.

4. Both the state and the belligerent are constrained in an IAC.

Table 9: Payoffs for constrained belligerent

The payoffs in this table combine the state payoffs of Table 5: Payoffs for constrained states

and Table 7: Payoffs for constrained insurgent. There is however one difference: the gain

awarded to compliance is estimated to be higher than for states. This has to do with the

belligerent’s status as a recognised formal participant in an IAC. The belligerent is assumed to

have a greater interest in future dealings, because of the ambition to over time participate as

an autonomous member of the international community in its own right.

68

Model 7: Sequential game for constrained belligerent

This increased interest does not in itself exert a greater pull towards compliance. In this

scenario too, the belligerent will have to communicate a signal of its intent to comply with the

agreement to his state counterpart. Otherwise the expectation to comply will not be sufficiently

supported and the most likely outcome remains mutual defection. The dynamics of legal

obligation and reciprocity seem to operate in a way similar to the dynamics of inter-state armed

conflicts. In the following paragraphs some preliminary conclusions will be linked to how they

could be improved upon in practice.

4.2.4 Practical implications

From the models it becomes apparent that the common conjectures of the intention to comply

are essential for establishing durable cooperation. Without the shared expectation the other will

comply, cooperation is highly unlikely in most scenarios. The dynamic requires AOG’s to signal

their intent to comply with IHL to opposing parties. Improving on the interactive pathways

between AOG’s and states could possibly lead to better compliance. The approach taken shows

that legal obligations of AOG’s to uphold humanitarian standards could provide both State- and

Non-State Actors with extra incentives to comply with humanitarian law. That is, only if the

AOG’s have an interest in future dealings. Any effort to improve IHL compliance should be

directed at providing AOG’s that have none, with an interest in future dealings. For AOG’s that

already have an interest in future dealings, any efforts to nurture this interest would lead to an

upward adjustment of payoffs for compliance, thus exert a stronger pull to compliance. This

leads me to think that the practice of recognizing AOG’s and their objectives bears the potential

of increasing compliance. However, this could yield perverse effects when AOG’s pursue goals

that are impossible to legitimise. This observation relates to the separation of jus in bello and

jus ad bellum. The neutrality of IHL to the causes of war is designed to constrain actors in any

armed conflict. Otherwise it could depend on the opinion of one sovereign entity which

opponents deserve protection and which do not. This could evidently lead to abuse and

69

discriminatory use of force. Touching the principle of equal application would engender

dangerous implications.140 But maybe there is a way to reconcile the principle of equal

application with increasing the possibilities for AOG’s to adhere to IHL obligations. Maybe the

international forum the ICRC is working on can install a think tank on this issue. Maybe the new

progress in soft law instruments between State- and Non-State Actors can contribute to

bridging the gap. Or maybe the concept of recognition of an AOG as an entity capable of

acquiring legal rights and duties should be thought separate from recognizing the cause of

armed conflict. This last observation is what proves the most promising for better IHL

compliance in my own humble opinion. This chapter will conclude with a third part on

enforcement of IHL and challenges for IHL in connection to AOG’s. The idea of increasing IHL

compliance of AOG’s through development of an international agreement stating their

obligations under IHL and human rights law to which the AOG’s can accede has been voiced.

Essential for this to work is that they would also be provided with some reciprocal rights,

increasing their chances to obtain their objectives in a peaceful, regulated manner. 141

140 A. ROBERTS, “The equal application of the laws of war: a principle under pressure”, IRRC 2008, 961-962. 141 N. BHUTA, “The role International Actors Other Than States can Play in the New World Order” in A.

CASSESE (Ed.), Realizing Utopia, p. 74.

70

4.3 ENFORCEMENT AND SUGGESTIONS

4.3.1 Enforcement

For most of this section I can simply refer to the section under 3.4. The challenges for IHL

compliance of a general sort apply to AOG’s a fortiori. Some aspects need to be nuanced in light

of AOG’s specifically. In contrast with the actor-centric approach of rights and duties under IHL,

the approach of enforcement displays a statist nature. This brings about several issues leading

to the unfortunate confrontation with a deficit in enforcement of humanitarian violations.142 The

requirements for IHL to apply to AOG’s take into account whether or not the AOG has an

identifiable authoritative body and whether they represent some form of governing force. This

serves the end of enforcing violations on a level equivalent to state responsibility. However,

articles 14 and 15 of the ILC articles on State Responsibility clarify the extent of state

responsibility in case of insurgency. If the uprising fails to secure victory, the government

against which the uprising was aimed, is not responsible for violations of the insurgency. The

successful insurgent campaign that manages to install a new government brings about

responsibility for both the previous and the new governments violations.143 In that case, under

the unlikely hypothesis that logistical and institutional features allow for it, the mechanism of

State-Responsibility might provide recourse. However, at the end of an armed conflict, the

victorious party will most likely be interested in sanctions for violations committed by the other

side. Victor’s justice prevents IHL enforcement to its full extent. The dichotomy between the

laws of war and the cause of war is not absolute as regards some aspects like this. For now it is

still unclear what the contours of an equivalent regime of state responsibility for AOG’s would

look like.144 For the other part of enforcement, the same mechanism of individual criminal

responsibility applies to individual operatives of AOG’s.

4.3.2 Challenges and possible areas of improvement

It goes without saying that the challenges discussed above for IHL compliance in general, also

go for armed conflicts with NSA’s participating in them. One item that might need to be

142 E. GRIFFIN, & B. ÇALI, “International Humanitarian Law” in B. ÇALI (Ed.), International Law For

International Relations, 252. 143 E. DAVID, Principes, 740-742. 144 C. RYNGAERT, “State Responsibility and Non-State Actors” in M. NOORTMANN et al. (Eds.), Non-State

Actors in International Law, Oxford, Hart, 2015, 163-177.

71

nuanced in this context relates to the agential problems due to the hierarchical command

structure of state armed forces. For AOG’s, whether this constitutes a problem or not, depends

on the extent to which they are in fact organised hierarchically. In general one might posit that

AOG’s will tend to be less organised than state armed forces, since this requires resources

AOG’s do not have apparent access to. This implies that the problems of noise and individual

violations are the more probable for armed conflicts with AOG participation. The better an AOG

is organised, the less frequent individual violations will occur. For fragmented AOG’s, lacking

strong command, it is more difficult to engage with IHL standards. The possibility for individual

violation increases if a general top-down policy, training and discipline is lacking.145

Concerning the cause of their noncompliant behaviour, a general lack of incentive resulting in

poor IHL observation on their part can be denoted. This observation has everything to do with

an apparent asymmetry in comparison with their state counterparts. The AOG’s feel

disadvantaged in military power, which they might want to compensate for by rejecting

limitations on warfare. When brought into a judiciary system that is not up to dealing with these

new actors, the perceived unequal treatment causes AOG’s to reject the enforcement

mechanisms in place. This decreases the value they award to the cost of violation, which will

lead to less incentives for compliance.146 More differences resulting from this asymmetry relate

to the fact that they are not institutions part of a domestic legal order of discipline and judicial

scrutiny contrary to state actors. Nor do they meet the criteria to benefit from protection under

IHL under the statute of combatant privilege, yet they are exposed to the full-fledged

consequences of breaching their obligations. This results in a lack of reciprocity leading to little

or no incentive to comply. This issue should primarily be addressed in order to do better. One

suggestion to overcome this hurdle lies in the idea of a modified combatant privilege for armed

opposition groups.147 Analogous to the immunity provided for by the combatant privilege under

the condition that IHL was fully observed, an immunity from prosecution under criminal law for

non-state combatants would provide incentive to comply with IHL. However, for this approach

to be successful there should also be an equivalent to the ius ad bellum that states are held to

in order to prevent encouraging armed violence. Melzer suggests the penalization of taking part

in armed conflict, even though the humanitarian standards were met, under a specific offence of

breach of public peace. This would at the same time amount to a more legitimate way in dealing

145 E. DECREY WARNER, “Engaging with Non-State Armed Groups: Lessons Learned”, 36th Round Table on

Current Issues of International Humanitarian Law, “Respecting IHL: Challenges and Perspectives”, Institute

of International Humanitarian Law San Remo, 5-7 September 2013, 2-5. 146 E. DECREY WARNER, “Engaging with Non-State Armed Groups: Lessons Learned”, 36th Round Table on

Current Issues of International Humanitarian Law, “Respecting IHL: Challenges and Perspectives”, Institute

of International Humanitarian Law San Remo, 5-7 September 2013, 2-5. 147 N. MELZER, “Bolstering the Protection of Civilians in Armed Conflict”, in A. CASSESE (Ed.), Realizing

Utopia, p. 515-517.

72

with non-state belligerents, compared to the approach of labelling them as “unlawful

combatants” or “terrorists” and justify use of force in less than optimal circumstances.

Another observation that stands to reason relates to violations committed by Non-State Actors

and reciprocal state action. Although Morrow’s work points to greater compliance by

democracies that are legally bound, even those state actors can be swayed to violate IHL norms

through reciprocal action.148 NSA’s that fail to recognise IHL standards, expose themselves to

the risk of strategic reciprocity. NSA’s adopting extremely violent strategies or acting under the

radar of open warfare, must create more incentives for states to defect from their legal

obligations if no rational alternative leads to valuable enough payoffs.

148 J. D. MORROW, “When Do States Follow the Laws of War?”, American Political Science Review, 2007, vol.

101., no. 3., 571.

73

74

5 Conclusion

5.1 EVALUATIVE REMARKS “[H]ow institutions can be designed to encourage states to

pursue their own self-interest in ways that facilitate long-term

cooperation over time (…) may also be a concern in international law,

yet it is rarely addressed so explicitly in the rationalist, strategic and

game-theoretic language and approach that dominates much of the

international relations literature on state compliance with international

norms and legal obligations.”149 - F. .B. Adamson & C. L. Sriram

The rather long quote on top of this page I chose because it manages to say in the vocabulary

of the subject what I wanted to achieve with this thesis: if nothing else, I wanted to convey the

message that IRT and IL should be brought closer together to establish a mutually beneficial

dynamic. Both disciplines can gain a lot from each other’s respective insights. I believe it to be

imperative for students of International Law to be educated on a multidisciplinary level. Law is a

societal construct, thus it cannot be thought separately from actual and historical international

relations, political science and social science in general. I believe our educational system would

benefit much from broadening the scope of possibilities for students to adopt insights from

other disciplines into their respective domains.

When consulting the bibliography the reader might notice a lack of references to jurisprudence.

The theoretical and interdisciplinary approach justifies this omission. The examples provided for

to link theory with practice are already referenced to in doctrinal sources. Other examples are

too recent phenomena to reflect final jurisprudence. This demonstrates my intention to

emphasise the actual relevance of the topic by choosing such examples.

On a final self-critical note, one might reproach the purely theoretical way this dissertation

tackles some issues. I have tried throughout the dissertation to justify this approach and

account for some of the difficulties engendered by the theoretical/practical dichotomy. Within its

limited scope this dissertation cannot nor should be anything more than an initiative to combine

different international disciplines and an exploration of an idea. Hopefully this dissertation

provides an argument to explore its potential further. Otherwise it is also limited in a couple of

149 F. B. ADAMSON & C. L. SRIRAM, “Perspectives on international law in international relations”, in B. ÇALI

(Ed.), International Law For International Relations, 36.

75

ways. For one, there is an enormous gap between theorizing on conduct in war and actually

waging one on the battlefield. Obviously the reduction of strategic decisions to some frames and

arrows in a dissertation cannot do honour by its object entirely. I have tried to bridge the gap

by providing illustrative models, referring to practice and pointing to the practical implications of

any inferences. I have treated these aspects as subordinate. My primary aim was to construe a

logical, structurally sound theoretical paradigm. As a result this dissertation could do with more

illustrations born from practice. The models I generated are also purely illustrative, thus any

inferences from them are highly theoretical and must be tested with actual data and further

research. Still I am convinced of the potential benefits theory can provide. The usefulness of

doctrinal contemplation on societal issues is under no debate as far as my opinion is concerned.

This dissertation has hopefully convinced the reader that RCT provides useful insights for

ameliorating IHL compliance.

5.2 CONCLUSION

To summarise the course this dissertation took: first, the subject and the methodology were

situated in the introduction. Different concepts and notions that would be applied throughout

the dissertation were discussed in chapter two, which concluded with a section on contemporary

challenges for IHL as to better situate the problem area of the subject. Chapter three began

with a word on the specific branch of IRT this dissertation was to implement. The underlying

assumptions of a Rational Choice framework for compliance on the state level were set forth in

the second section. In the third section the dynamic and body of IHL norms was then analysed

and illustrated through the expounded paradigm. A fourth section served to discuss

enforcement at the state level and possible areas of improvement. The fourth chapter

transposed the framework generated in chapter three to compliance of AOG’s. The first section

served to accurately demarcate the contours of this endeavour and testing the underlying

assumptions. In the second section the dynamic of IHL compliance betweens states and AOG’s

was analysed and illustrated through the paradigm, in order to indicate some conclusions linked

to their practical implications. The third section was dedicated to highlighting some nuances

regarding enforcement and areas of improvement regarding compliance of AOG’s in particular.

A short fifth chapter served to put some aspects of this dissertation into perspective.

One main conclusion this dissertation leads to draw is that IL matters. More specifically IHL can

effectively constrain behaviour, both of states and AOG’s. This statement answers the first

question this dissertation departed from. As to the second question, the approach of RCT for

analyzing incentives to comply has proven promising. IHL compliance would benefit from

creating incentives for AOG’s to partake in the international community. Moreover, existing

76

incentives must be strengthened for fewer violations to occur. Emancipation of AOG’s is a

promising way of achieving this. For one, ownership of their obligations incites compliance. This

could be promising for a new approach to enforcement of IHL vis-à-vis AOG’s. And finally, this

would serve to support the credibility of any commitment AOG’s make towards IHL. This is

critical in order to establish stable State-AOG relations that rely on sustainable cooperation. As

an international institution in the “IRT sense” of the term, humanitarian law is still relatively

young and has a myriad of battles to conquer. The design of the institute could be adjusted by

addressing some of the issues that surfaced in this dissertation. Hopefully over time there will

be no more use for IHL entirely. How nice would it be if all international quarrels would be

settled by a good game of Risk.

77

78

Bibliography

LEGISLATION Common Article 2 of the Geneva Conventions, 1949.

Common Article 3 of the Geneva Conventions, 1949.

Article 1, §4 of the Additional Protocol (I) to the Geneva Conventions, 1977.

Article 90 of the Additional Protocol (I) to the Geneva Conventions,1977.

Additional Protocol (II) to the Geneva Conventions, 1977.

Article 19 Hague Convention for the Protection of Cultural Property, 1954.

Article 22 of the Second Hague Protocol for the Protection of Cultural Property.

Article 4 Statute of the International Criminal Tribunal for Rwanda, 1994.

Article 3 of the Statute for the Special Court of Sierra Leone, 2002.

Second Protocol with the Convention Prohibiting Certain Conventional Weapons (CCW) of 1980,

as amended in 1996.

Convention prohibiting Certain Conventional Weapons (CCW), amended Article 1, 2001.

CCW Protocol (I) on Non-Detectable Fragments, 1980.

CCW Protocol (II) prohibiting Mines, Booby-Traps and Other Devices, 1980.

CCW Protocol (III) prohibiting Incendiary Weapons, 1980.

CCW Protocol (IV) on Blinding Laser Weapons, 1995.

CCW Protocol (V) on Explosive Remnants of War, 2003.

79

Article 8 of the Statute of the International Criminal Court, 1998.

Articles 14, 15 and 41, (2) of the ILC Articles on State Responsibility.

Articles 34-36 of the Vienna Convention on the Law of Treaties, 1969.

Resolution 2209 of the United Nations Security Council (6 March 2015), UN Doc. S/RES/2209

(2015).

Resolution 1 – Strengthening legal protection for victims of armed conflict, 31st international

Conference 2011, ICRC, 1 December 2011.

Concluding Report of the 32nd International Conference of the Red Cross and the Red Crescent,

Accessed online at: https://www.icrc.org/eng/resources/documents/misc/57jmbh.htm;

Report of the Independent International Commission of Inquiry on the Syrian Arab Republic,

Human Rights Council, A/HRC/23/58, 4 June 2013.

Report of the Independent International Commission of Inquiry on the Syrian Arab Republic,

Human Rights Council, A/HRC/31/68, 11 February 2016.

DAVID, D, TULKENS, F. & VANDERMEERSCH, D. (Eds.), Code de droit international humanitair,

Brussels, Bruylant, 2013, 1042 p.

DAVID, E, & VAN ASSCHE, C., Code de droit international public, Brussels, Bruylant, 2011, 1370

p.

DOCTRINAL SOURCES

Books

BEARDSWORTH, R., Cosmopolitanism and International Relations Theory, 2011, Cambridge,

Polity Press, 272 p.

BILLIET, J. en WAEGE, H. (Eds.), Een samenleving onderzocht. Methoden van sociaal-

wetenschappelijk onderzoek, Antwerp, De Boeck, 2006, 390 p.

80

BURGSTALLER, M., Theories of Compliance with International Law, Leiden, Martinus Nijhoff,

2005, 244 p.

BUTLER, W. E. (Ed.), Control over Compliance with International Law, Dordrecht, Martinus

Nijhoff, 1991, 209 p.

BYERS, M., Custom, Power and the Power of Rules: International Relations and Customary

International Law, Cambridge, Cambridge University Press, 2001, 250 p.

CARLSNAES, W., RISSE, T. & SIMMONS, B. A. (Eds.), Handbook of International Relations,

London, Sage Publications, 2002, 904 p.

CARMICHAEL, F., A guide to game theory, Harlow, Pearson Education Ltd., 2005, 286 p.

CHENG, T.-H., When International Law Works: Realistic idealism after 9/11 and the global

recession, New York, Oxford University Press, 2012, 341 p.

CLARK, I., Globalization and International Relations Theory, Oxford, Oxford University Press,

1999, 197 p.

CORTEN, O., Méthodologie de droit international public, Brussels, Editions de L’Université de

Bruxelles, 2009, 291 p.

DAVID, E., Principes de Droit des Conflits Armés, Brussels, Bruylant, 2012, 1150 p.

DETTTER, I., The Law of War, Cambridge, Cambridge University Press, 2000, 516 p.

FALK, R. A., The Costs of War: International Law, the UN and World Order after Iraq, New York,

Routledge, 2008, 203 p.

GOLDSMITH, J. & POSNER, E., The Limits of International Law, New York, Oxford University

Press, 2005, p. 272.

GRIFFITHS, M., ROACH, S. C., & SOLOMON, M. S., Fifty Key Thinkers in International Relations,

2009, Oxon, Routledge, 404 p.

GUZMAN, A. T., How International Law Works: A Rational Choice Theory, New York, Oxford

University Press, 2008, 272 p.

HENKIN, L., How Nations Behave, New York, Columbia University Press, 1979, 400 p.

81

JØRGENSEN, K. E., International Relations Theory: A New Introduction, Hampshire, Palgrave

Macmillan, 272 p.

KAHNEMAN, D., Thinking Fast and Slow, London, Penguin Books, 2011, 499 p.

KALDOR, M., New and Old Wars: Organized Violence in a Global Era, Stanford, Stanford

University Press, 2012, p.

KAURIN, P. M., “Enemy Combatant” in CHATTERJEE, D. K., (Ed.), Encyclopedia of Global

Justice, Salt Lake City, Springer, 2012, v° “Enemy Combatant”, 1200 p.

KOSKENNIEMMI, M., From Apology to Utopia: The structure of International Legal Argument,

Cambridge, Cambridge University Press, 2006, 683 p.

KRATOCHWIL, F. V., “Rules, Norms, and Decisions: On the Conditions of Practical and Legal

Reasoning” in International Relations and Domestic Affairs, Cambridge University Press, 1989,

317 p.

KRIEGER, H. (Ed.), Inducing Compliance with International Humanitarian Law: Lessons from the

African Great Lakes Region, Cambridge, Cambridge University Press, 2015, 557 p.

KUMAR, D., Islamophobia and the politics of Empire, Chicago, Haymarket Books, 2012, 238 p.

KYDD, A. H., Trust and Mistrust in International Relations, Princeton, Princeton University Press,

2005, 284 p.

KYDD, A. H., International Relations Theory: the Game-theoretic approach, Cambridge,

Cambridge University Press, 2015, 228 p.

LUBELL, N., Extraterritorial Use of Force Against Non-State Actors, Oxford, Oxford University

Press, 2010, 320 p.

MAY, L., Limiting Leviathan: Hobbes on law and international affairs, Oxford, Oxford university

press, 2013, 260 p.

MAY, L., FORCEHIMES, A. T., (Eds.), Morality, Just post Bellum and International Law,

Cambridge, Cambridge University Press, 2014, 282 p.

82

MAZOWER, M., “Chapter 3: The Empire of Law”, in Governing the World: The History of an

Idea, New York, The Penguin Press, 2012, p. 65-93.

MCDONALD, P. J., The invisible hand of peace: capitalism, The War Machine and International

Relations Theory, New York, Cambridge University Press, 2009, 338 p.

MORROW, J. D., Order within Anarchy: The Laws of War as an International Institution, New

York, Cambridge University press, 2014, 354 p.

OHLIN, J. D., The Assault on International Law, New York, Oxford University Press, 2015, 289 p.

OTTO, R., Targeted Killings and International Law. With Special Regard to Human Rights and

International Humanitarian Law, Heidelberg, Springer, 2011, 661 p.

PAUST, J. L. et al., International Criminal Law, Durham NC, Carolina Academic Press, 200, 335.

RAJKOVIC, N. M., The Politics of International Law and Compliance: Serbia, Croatia and The

Hague Tribunal, Oxon, Routledge, 2012, 213 p.

SAXON, D. (Ed.), International Humanitarian Law and the Changing Technology of War, Leiden,

Martinus Nijhoff, 2013, 358 p.

SCHELLING, T. C., The Strategy of Conflict, Cambridge, Harvard University Press, 1980, 309 p.

SIMMONS, B. A. & STEINBERG, R. H. (eds.), International Law and International Relations,

2006, Cambridge, Cambridge University Press, 776 p.

SINGER, P. W., Corporate Warriors: The Rise of the Privatised Military Industry, New York,

Cornell University Press, 2007, 348.

SMIS, S., JANSSENS, C., MIRGAUX, S., & VAN LAETHEM, K., Handboek Mensenrechten: De

internationale bescherming van de rechten van de mens, Antwerp, Intersentia, 2002, 660.

SOLIS, G. D., The law of Armed Conflict: International Humanitarian Law in War, Cambridge,

Cambridge University Press, 2010, 659 p.

STEINBERG, G. M., & HERZBERG, A., Best Practices for Human Rights and Humanitarian Fact-

Finding”, Leiden, Brill, 2012, 212 p.

83

WALT, S. M., The origins of alliance, 1978, Cornell University Press, 110-115.

ZEGVELD, L., Accountability of Armed Opposition Groups in International Law, Cambridge,

Cambridge University Press, 2002, 260 p.

Contributions in Collective Doctrinal Works

ADAMSON, F. B., & SRIRAM, C. L., “Perspectives on international law in international relations”,

in ÇALI, B. (Ed.), International Law For International Relations, New York, Oxford University

Press, 2010, 25-46.

BENNOUNA, M., “How to Cope with the Proliferations of International Courts and Coordinate

Their Action” in CASSESE, A. (Ed.), Realizing Utopia: The Future of International Law, Oxford,

Oxford University Press, 2012, p. 287-296.

BHUTA, N., “The role International Actors Other Than States can Play in the New World Order”

in CASSESE, A. (Ed.), Realizing Utopia: The Future of International Law, Oxford, Oxford

University Press, 2012, p. 61-75.

CASSESSE, A., “Towards a Moderate Monism: Could International Rules Eventually Acquire the

Force to Invalidate Inconsistent National Laws?” in CASSESE, A. (Ed.), Realizing Utopia: The

Future of International Law, Oxford, Oxford University Press, 2012, p. 185-199.

CASSESSE, A., “The International Court of Justice: it is High Time to Restyle the Respected Old

Lady” in CASSESE, A. (Ed.), Realizing Utopia: The Future of International Law, Oxford, Oxford

University Press, 2012, p. 239-249.

CASSESSE, A., “Fostering Increased Conformity with International Standards: Monitoring and

Institutional Fact-Finding”, in CASSESE, A. (Ed.), Realizing Utopia: The Future of International

Law, Oxford, Oxford University Press, 2012, p. 295-303.

CASSESSE, A., “Should Rebels be Treated as Criminals? Some Modest Proposals for Rendering

Internal Armed Conflicts Less Inhumane” in CASSESE, A. (Ed.), Realizing Utopia: The Future of

International Law, Oxford, Oxford University Press, 2012, p. 519-524.

84

CHAYES, A. & HANDLER CHAYES, A., “On Compliance” in SIMMONS, B. A. & STEINBERG, R. H.

(Eds.), International Law and International Relations, Cambridge, Cambridge University Press,

2006, p. 65-91.

CONDORELLI, L., “Customary International Law: The Yesterday, Today, and Tomorrow of

General International Law” in CASSESE, A. (Ed.), Realizing Utopia: The Future of International

Law, Oxford, Oxford University Press, 2012, p. 147-157.

CLAPHAM, A., “Overseeing Human Rights Compliance”, in CASSESE, A. (Ed.), Realizing Utopia:

The Future of International Law, Oxford, Oxford University Press, 2012, p. 318-325.

D’ASPREMONT, J., “International Law-Making by Non-State Actors: Changing the Model or

Putting the Phenomenon into Perspective”, in NOORTMANN, M. & RYNGAERT, C. (Eds.), Non-

State Actor Dynamics in International Law: From Law-Takers to Law-Makers, Farnham,

Ashgate, 2010, 195-202.

DUPUY, P.-M., “The Deficiencies of the Law of State Responsibility Relating to Breaches of

‘Obligations Owed to the International Community as a Whole’: Suggestions for Avoiding the

Obsolescence of Aggravated Responsibility” in CASSESE, A. (Ed.), Realizing Utopia: The Future

of International Law, Oxford, Oxford University Press, 2012, p. 210-226.

GAETA, P., “Immunity of States and State Officials: A Major Stumbling Block to Judicial

Scrutiny?” in CASSESE, A. (Ed.), Realizing Utopia: The Future of International Law, Oxford,

Oxford University Press, 2012, p. 227-238.

GAUBATZ, K. T., “Democratic States and Commitment in International Relations” in SIMMONS,

B. A. & STEINBERG, R. H. (Eds.), International Law and International Relations, Cambridge,

Cambridge University Press, 2006, p. 43- 64.

GRIFFIN, E. & ÇALI, B., “International Humanitarian Law” in ÇALI, B. (Ed.), International Law

For International Relations, New York, Oxford University Press, 2010, 234-257.

HERMAN, J., “International Law and humanitarian space in the twenty-first century: challenged

relationships” in ZWITTER, A., LAMONT, CH. K., HEINTZE, H.-J., & HERMAN, J. (Eds.),

Humanitarian Action: Global, Regional and Domestic Legal Responses, Cambridge, Cambridge

University Press, 2015, p. 11-32.

KELSEY, J. T. G., “Hacking into International Humanitarian Law: Principles of Distinction and

Neutrality in the Age of Cyber Warfare”, Mich. L. Review 2008, p. 1427-1452.

85

KORNPROBST, M., “Non-State Actors in International Relations” in NOORTMANN, M., REINISCH,

A., RYNGAERT, C. (Eds.), Non-State Actors in International Law, Oxford, Hart, 2015, p. 295-

322.

KRASNER, S. D., “Structural Causes and Regime Consequences: Regimes as Intervening

Variables” in SIMMONS, B. A. & STEINBERG, R. H. (Eds.), International Law and International

Relations, 2006, Cambridge, Cambridge University Press, p. 3-17.

KUIJT, E. E., “A humanitarian crisis: reframing the legal framework on humanitarian assistance”

in ZWITTER, A., LAMONT, CH. K., HEINTZE, H.-J. & HERMAN, J., (Eds.), Humanitarian Action:

Global, Regional and Domestic Legal Responses, Cambridge, Cambridge University Press, 2015,

p. 54-80.

MELZER, N., “Bolstering the Protection of Civilians in Armed Conflict” in CASSESE, A. (Ed.),

Realizing Utopia: The Future of International Law, Oxford, Oxford University Press, 2012, p.

318-325.

NOORTMANN, M., “Understanding Non-State Actors in the Contemporary World Society:

Transcending the International, Mainstreaming the Transnational of Bringing the Participants

Back In?”, in NOORTMANN, M. & RYNGAERT, C. (Eds.), Non-State Actor Dynamics in

International Law: From Law-Takers to Law-Makers, Farnham, Ashgate, 2010, 153-170.

NIJMAN, J. E., “Non-State Actors and the International Rule of Law: Revisiting the ‘Realist

Theory’ of International Legal Personality in NOORTMANN, M. & RYNGAERT, C. (Eds.), Non-

State Actor Dynamics in International Law: From Law-Takers to Law-Makers, Farnham,

Ashgate, 2010, 91-124.

PAUST, J., “Armed Opposition Groups”, in M. NOORTMANN, A. REINISCH & C. RYNGAERT

(Eds.), Non-State Actors in International Law, Oxford, Hart, 2015, 273-292.

RONZITTI, N., “Modern Means of Warfare: The Need to Rely upon International Humanitarian

Law, Disarmament, and Non-Proliferation Law to Achieve a Decent Regulation of Weapons” in

CASSESE, A. (Ed.), Realizing Utopia: The Future of International Law, Oxford, Oxford University

Press, 2012, p. 553-570.

RYNGAERT, C., “State Responsibility and Non-State Actors”, in NOORTMANN, M., REINISCH, A.,

RYNGAERT, C. (Eds.), Non-State Actors in International Law, Oxford, Hart, 2015, 163-182.

86

SCHABAS, W., “The International Criminal Court: Struggling to Find its Way” in CASSESE, A.

(ed.), Realizing Utopia: The Future of International Law, Oxford, Oxford University Press, 2012,

p. 250-260.

SCHEININ, M., “Monitoring Human Rights Obligations and The Fight Against Terrorism: Whose

Obligations? And Monitored How?” in ALFREDDSON, J. GRIMHEDEN, J., RAMCHARAN, B.G.,

ZAYAS, A., (Eds.), International Human Rights Monitoring Mechanisms. Essays in Honour of

Jakob Th. Möller 2nd Revised Edition, Brill, 2009, p. 407–420.

SHANY, Y., “Should the Implementation of International Rules by Domestic Courts be

Bolstered?” in CASSESE, A. (Ed.), Realizing Utopia: The Future of International Law, Oxford,

Oxford University Press, 2012, p. 200-209.

SINCLAIR, A., International Relations Theory and International Law: A critical approach,

Cambridge, Cambridge University Press, 2010, 201 p.

SIVAKUMARAN, S., “How to Improve upon the Faulty Legal Regime of Internal Armed Conflicts”,

in CASSESE, A. (Ed.), Realizing Utopia: The Future of International Law, Oxford, Oxford

University Press, 2012, p. 525-537.

SIVAKUMARAN, S., “Implementing humanitarian norms through non-State armed groups”, in

KRIEGER, H. (Ed.), Inducing Compliance with International Humanitarian Law: Lessons from the

African Great Lakes Region, Cambridge, Cambridge University Press, 2015, 125-146.

SUN TZU, S., The Art of War, Courrier Corporation, 2002, 52 p.

TALMON, S., “The Duty not to “recognize as lawful” a Situation created by the Illegal Use of

Force or other Serious Breaches of a Jus Cogens Obligation: an Obligation without Real

Substance?” in TOMUSCHAT, C., & THOUVENIN, J.-M. (Eds.), The Fundamental Rules of the

International Legal Order: Jus cogens and Obligations Erga Omnes, Leiden, Martinus Nijhoff, 99-

125.

THÜRER, D., “International Law: Theory, Practice, Context” in Collected Courses of the Hague

Academy of International Law, The Hague, Hague Academy of International Law, 2011, 504 p.

VAROGLU, A. K., & UYAR, M., “The impact of asymmetrical warfare on the military profession

and structure: lessons learned from the ottoman military” in CAFORIO, G., KÜMMEL, G.,

PUERKAYASTHA, B., (Eds.), Armed Forces and Conflict Resolution: Sociological Perspectives,

Emerald Group Publishing, 2008, 411 p.

87

Articles

AXELROD, R. & KEOHANE, R. O., “Achieving Cooperation under Anarchy: Strategies and

Institutions”, World Politics 1985, vol. 38, no. 1, p. 226-254.

BAXTER, R. R., “Forces for Compliance with the Laws of War”, ASIL Proceedings 1964, p. 82-88.

BUCHANAN, A. & KEOHANE, R., “Towards a Drone Accountability Regime”, Ethics and

International Affairs 2015, p. 15-37.

BRADFORD, W., “In the Minds of Men: a Theory of Compliance with the Laws of War”, Arizona

State Law Journal 2004, 1243-1439.

CLAPHAM, A., “Human Rights Obligations of Non-State Actors in Conflict Situations”, IRRC 2006,

p. 491-523.

CLAPHAM, A., “The Regulations of Private Military and Security Contractors: Introductory

Remarks”, ASIL Proceedings 2013, 199-200.

ENGLEHART, N. A., “Non-state Armed Groups as a Threat to Global Security: What Threat,

Whose Security?, Journal of Global Security Studies 2016, 1.

GREENWOOD, CH., “The Concept of War in Modern International Law”, ICLQ 1987, p. 283-306.

HENCKAERTS, J. M., “Study on customary international humanitarian law: A contribution to the

understanding and respect for the rule of law in armed conflict”, IRRC 2005, 175-212.

HENKIN, L., “How are nations behaving?”, ASIL proceedings 2002, p. 205-218.

KOH, H., “Why Do Nations Obey International Law?”, The Yale Law Journal 1997, p. 2599-2639.

LEGGIERO, K., “Countering ISIS Recruitment in Western Nations”, Journal of Political Risk 2015,

accessed online at: http://www.jpolrisk.com/countering-western-recruitment-of-isis-fighters/

MEARSHEIMER, J., “The False Promise of International Institutions”, International Security,

1994-95, p. 5-49.

88

MERON, T., “On the Inadequate Reach of humanitarian and Human Rights Law and the Need for

a New Instrument”, AJIL 1983, p. 589-606.

MERON, T., “The Humanization of Humanitarian Law”, AJIL 2000, p. 239-278.

MORROW, J., “When Do States Follow the Laws of War?”, American Political Science Review,

2007, 559-572.

NEWTON, M. A., “Flying into the Future: Drone Warfare and the Changing Face of Humanitarian

Law” Keynote Address at the 2010 Sutton Colloquium, Denv. J. International. L Pol’y 2010, p.

601-614.

OKIMOTO, K., “Violations of International Humanitarian Law by United Nations Forces and their

Legal Consequences”, Yearbook of International Humanitarian Law 2003, 199-236.

RICHEMOND-BARAK, D., “Applicability and Application of the Laws of War to Modern Conflict”,

Florida Journal of International Law 2012, 327-357.

ROBERTS, A., “The equal application of the laws of war: a principle under pressure”, IRRC 2008,

931-962.

RODENHÄUSER, T., “International legal obligations of armed opposition groups in Syria”,

International Review of Law 2015, 16 p. accessed online at http://dx.doi.org/10.5339/irl.2015.2

SIVAKUMARAN, S., “Binding Armed Opposition Groups”, International and Comparative Law

Quarterly 2006, p. 369-394.

STEWART, J. G., “Towards a single definition of armed conflict in International Humanitarian

Law: A critique of internationalized armed conflict”, IRRC 2003, 313-350.

TÉSON, F. R., “The Kantian Theory of International Law”, Columbia Law Review 1992, p. 53-

102.

TRACHTMAN, J. P., “International law and Domestic Political Coalitions: The Grand Theory of

Compliance with International Law”, Chicago Journal of International Law 2010, p. 127-158.

VAN AAKEN, A., “Behavioral International Law and Economics”, Harvard International Law

Journal 2014, p. 421-481.

89

VERDIER, P.-H. & VOETEN, E., “Precedent, compliance and change in customary international

law: an explanatory theory”, AJIL 2014, p. 389-434.

ZOLI, C., SCHNEIDER, E., & SCHUSTER, C., “Armed conflict and Compliance in Muslim States,

1947-2014: Does Conflict Look Different under International Humanitarian Law?”, N.C.J. Int’l L.

& Com. Reg. 2015, 679-738.

Other

AL-MAAGHEED, M. A., “The Arab Spring and the Observance of IHL in Recent Conflicts”, 36th

Round Table on Current Issues of International Humanitarian Law, 2013, 14 p.

DAM-DE JONG, D., “Armed Opposition Groups and the Right to Exercise Control over Public

Natural Resources: A Legal Analysis of the Cases of Libya and Syria”, Neth Int Law Rev 2015, 3-

34.

DECREY WARNER, E., “Engaging with Non-State Armed Groups: Lessons Learned”, 36th Round

Table on Current Issues of International Humanitarian Law, “Respecting IHL: Challenges and

Perspectives”, Institute of International Humanitarian Law San Remo, 5-7 September 2013, 5

p.; online source: www.iihl.org/iihl/Documents/DECREY-WARNER.pdf, accessed 24 November

2014.

DEL PONTE, C., “The Observance of IHL in Recent Conflicts: The Case of Syria”, 36th Round

Table on Current Issues of International Humanitarian Law, 2013, 4 p.

FOSTER HALABI, S., “Traditions of Belligerent Recognition: the Libyan Intervention in Historical

and Theoretical Context”, Am. U. Int’l L Rev. 2012, 321-389.

GRAHAM-HARRISON, E., “Taliban fears over young recruits attracted to Isis in Afghanistan”,

The Guardian 2015, accessed online at:

http://www.theguardian.com/world/2015/may/07/taliban-young-recruits-isis-afghanistan-

jihadis-islamic-state.

HOWSE, R. & TEITEL, R., “Beyond compliance: Rethinking Why International Law Really

Matters” in New York University of Public Law and Legal Theory Working Papers, Paper 174,

New York, New York University, 2010, 29 p.

90

MAURER, P., Strengthening compliance with international humanitarian law, speech at the 4th

meeting of States on strengthening compliance with humanitarian law, Geneva, 23 April 2015,

https://www.icrc.org/en/document/fourth-meeting-states-strengthening-compliance-

international-humanitarian-law.

PROCTER, P. (Ed.), Longman Dictionary of Contemporary English, Harlow, Longman, 1978;

online version at www.longmandictionariesonline.com, v° “compliance”.

RICHEMONDE-BARAK, D., The 'Informal Law of War', working paper for the WBZ Berlin

Colloquium on Private Ordering and Public Authority, 2013, 25 p.

https://www.wzb.eu/sites/default/files/u32/richemond-

barak__the_informal_law_of_war_berlin.pdf;

SCHMITT, E. & SHANKER, T., “US Weighed Use of Cyberattacks to Weaken Libya”, NY Times

2011 (Oct. 17), accessed online at: http://www.nytimes.com/2011/10/18/world/africa/cyber-

warfare-against-libya-was-debated-by-us.html?_r=0

TALMON, S., “Recognition of Opposition Groups as the Legitimate Representative of a People”,

Chinese JIL 2013, 219-253.

X., International Non-Governmental Organisations Accountability Charter, 2005,

www.ingoaccountabilitycharter.org

X., Respect for International Humanitarian Law, in Handbooks for Parliamentarians, ICRC,

Geneva, 1999, 104 p

91

Appendix

SCHEMATIC OVERVIEW OF IHL SCOPE OF APPLICATION

92

DETAILED JUSTIFICATION OF TABLES AND MODELS

INCLUDED

Table 1: A simple coordination game

Each player gains 5 by complying and gains 10 for unilateral defection. This would result in a

loss of -5 for the unilateral compliant side. Mutual defection would result in an even greater loss

of -15.

In this simple coordination game, both players’ dominant strategy is to pursue the payoff of 10

by defecting. This payoff would only be realised, if the counterpart complies. Since that

equilibrium results in -5 for Actor 2, Actor 1 can safely expect Actor 2 not to comply. Should

neither of them comply, both are worse off, resulting in a payoff of -15 for each. The best

possible outcome is thus for each to comply, since it is better to settle for a payoff of 5, than

risking the payoff of -15. If the actors coordinate their action this is the most likely outcome.

Table 2: A one-shot Prisoner’s Dilemma

Each actor gains 5 by complying, 10 from unilateral defection and -5 from unilateral compliance.

The difference is that the outcome for mutual defection result in a positive outcome of 3 for

each.

In the difficult coordination game of a Prisoner’s Dilemma the most likely outcome in a one-shot

game is for both parties to defect. Defection is the dominant strategy for each, since it would

generate a payoff of 10, should the other comply. However, since unilateral compliance leaves

93

the compliant actor with a payoff of -5, neither of both will want to take the risk of complying. If

one of them can safely expect the other’s strategy to be ‘comply’, this increases the incentive of

the other to defect, securing the maximal payoff of 10. Therefore mutual defection is the most

likely outcome. Actors prefer settling for a payoff of 3 to running the risk of being left with -5:

the strategy ‘defect’ provides the opportunity to exploit the other’s compliance, and secures

protection against the other’s exploitation. At the same time this strategy secures the smallest

possible payoff for the counterpart.

Table 3: Assurance game: trustworthy players

In the assurance game both players can trust the other to comply since the equilibrium

(comply, comply) is the dominant strategy for both. Should Actor 1 comply, he can trust Actor 2

will comply as well, since that generates him a payoff of 5 over a payoff of 4 under the strategy

of defection. Should Actor 1 defect, so will Actor 2 (choosing a payoff of 3 over -5), yet this

leaves both parties worse off. (Comply, comply) is both actors’ best strategy.

Model 1: Tree model for a sequential game

As each actor has two moves to choose from this situation results in four hypothetical

situations:

- The inferior side chose to defect,

94

- The superior side has to choose whether to retaliate or not. If it does not it will suffer

a loss due to the other side’s defection and the other side will have obtained a high payoff. This

scenario is not likely to occur.

- If the superior side does decide to reciprocate/retaliate, the inferior side will most

likely be crushed. In this scenario both endure transactional costs, but the inferior side will also

face a high cost of retaliation. This scenario is unlikely, as it would not be rational for the

inferior state to expose itself to violation of the other side and it should most certainly expect

the superior side to reciprocate/retaliate.

- The inferior side chose to comply; the superior side has to decide between two moves: comply

as well or violate anyway.

- Should the superior side decide to violate, this would only be rational if the payoff

compensates for the transactional cost of violation. It is hard to imagine what rational incentive

this side would have, since as the superior power his threats will definitely deter violation from

the inferior side.

- This leaves us with the fourth and most likely outcome: the inferior side complies out

of fear for a credible threat of reciprocation/retaliation. The superior side complies, as it has no

incentive to defect as long as the other side complies. Both gain by cooperation, their joint total

payoff is maximal in this scenario.

Table 4: Payoffs for unconstrained states

The payoffs represented here are the independent payoffs (IPO) of military strategy. The same

dynamic works here as did in model 2: defect is the dominant strategy for each, so in order to

prevent exploitation by the other, (defect, defect) would be the most likely outcome in a one-

shot game. There is no guarantee the other would comply, even though this would constitute a

mutual best outcome, since the payoff for unilateral defection is a lot higher. However, the

game is repeated. Since there is no agreement on cooperation in place, this would likely

represent a war of attrition.

95

Model 2: Sequential game for unconstrained states

In the sequential game model the asymmetry in the hypothesis of retaliation becomes apparent

too. Should the inferior state comply, it would expose himself to the opportunity of exploitation

by the superior state. Hence, the inferior state would rather settle for a payoff of 10, minimizing

the superior state’s possible outcome. (Chances of the superior state not retaliating are slim, as

this would engender a payoff of -5).

Table 5: Payoffs for constrained states

In this table of payoffs the impact of treaty obligation becomes clear in the dependent payoffs

(DPO). These are the gains or costs related to the “three R’s”. Compliance is represented as an

increased payoff of 5, due to reputational gains and the omission of the transaction cost of

violation. Defection is represented in a cost of -10, since the transaction cost incurred will

probably be greater than the reputational cost of defection. State 1 and State 2 have been

awarded different estimations of the outcome for (defect, defect) in order to represent an

element of asymmetry regarding military power.

96

Model 3: Sequential game for constrained states

In the sequential game the total payoffs are indicated for each outcome. The legal obligation

clearly results in a pull towards compliance: the dominant strategy for each remains defection,

but if they expect each other to defect, they would be mutually worse off in the equilibrium

(defect, retaliate). It seems that, should the inferior state comply, the superior state could

exploit his compliance by defecting anyway. The dynamic of repetition prevents this from

happening. Should the superior state choose to defect, the inferior state will return the favour

at the next round. That is called the Tit-for-Tat strategy in game theoretic terms. Over time the

added payoffs of (comply, comply) will outweigh the difference of what payoffs the smaller

gains from (defect, retaliate). Mutual compliance results in the greatest joint payoff.

Table 6: Payoffs for unconstrained insurgent

The payoffs for the constrained state are based on the payoffs for State 2 in Table 5. Since

there was already an element of asymmetry represented in those payoffs, the estimations will

not vary too much in relation with an insurgent AOG. The estimations for the insurgent are

lower in each of the outcomes involving defection, due to the fact that it does not represent an

equal sovereign entity at this point. Where unilateral defection by State 1 is estimated at a

positive IPO of 40, the same IPO is estimated at half (20), since the AOG is estimated to do only

“half as much damage” to State 2. Where mutual defection by State 1 is estimated at a positive

97

IPO of 10, the same IPO for the AOG is estimated at 0. Due to its lesser capacity to damage his

opponent the gain from defection and the cost of defection are thought to cancel each other

out. The outcome for (comply, comply) is estimated at an equal value of 15, since one might

adduce they both have the same relative gain from compliance as regards their independent

objective payoff.

Model 4: Sequential game for unconstrained insurgent

In the sequential model the total payoffs reflect the aspect of asymmetry well, compared to an

inter-state armed conflict. Should the insurgent comply, it exposes him to the risk of

exploitation as the State can opt for a payoff of 30 instead of 20. The dominant strategy for the

insurgent is thus to defect, since it will minimalise the State’s payoff to either 0 or 3. The State

will then play “defect” as well, since it prefers 3 to 0. The equilibrium reached will be (defect,

defect) because there is no agreement in place that would allow the insurgent to expect

compliance from the state, should the insurgent comply first.

Table 7: Payoffs for constrained insurgent

The payoffs for the constrained State are again taken from Table 5. The independent payoffs of

the Insurgent are the same as in Table 6. The gain that is awarded to compliance has a value of

98

5, same as for constrained states. The Insurgent’s cost for defection is valued at -15, whereas

the same cost for the State is estimated at -10. This represents the idea that due to the

asymmetrical resources the Insurgent would suffer a greater loss from unilateral or reciprocal

State defection, since the State can do more damage.

Model 5: Sequential game for constrained insurgent

The effect of legal obligation for the total payoffs reflects the increased potential gain of

compliance and the increased cost of defection. The increased cost of defection represents the

difference that parties had to first incur a transaction cost to provide the other with a credible

commitment to upholding the standard. Imagine the AOG had to first negotiate an agreement

with the State; they both would have to “invest” resources in reaching the resulting agreement.

Hence the AOG has more to lose by defection, because it had invested in cooperation. The same

dynamic is not reflected in the incentive structure for states, because if the state were

constrained in this conflict situation, it would have been under any armed conflict situation. The

State had to incur those transaction costs when entering into the treaty, but it did so

independent from its relation to the AOG or the conflict. The AOG only committed because its

position in the armed conflict demanded or recommended it. The changed incentives have no

direct impact on the equilibrium. The common conjecture that the State will comply if the AOG

does, as a result of the agreement in place, however, will push the Insurgent towards (comply,

comply).

99

Table 8: Payoffs for unconstrained belligerent

The payoffs for the unconstrained belligerent are based on the payoffs for the unconstrained

Insurgent in Table 6. The payoffs for the unconstrained State are based on those for State 2 in

Table 4. This situation represents the IAC between a state and a recognised belligerent. In this

situation the state has the right to reciprocally suspend its obligations and is thus ‘temporarily’

unconstrained in relation to the AOG that does not ‘accept and apply IHL’.

Model 6: Sequential game for unconstrained belligerent

The effect this dynamic has is very similar to the dynamic represented in Model 2., with the sole

difference in the estimated payoffs for defection for the belligerent here. Due to asymmetry

they are estimated lower than for a state, even more so than those for the inferior State 1. This

does not affect which equilibrium will be played. The difference is the AOG will have to settle for

a payoff of 0 for (defect, defect) whereas inferior State 1 had to settle for 10.

100

Table 9: Payoffs for constrained belligerent

The payoffs for this incentive structure are based on Table 7. The sole difference is that the

belligerent AOG has more to gain from compliance than an insurgent. This represents the AOG’s

wish to be recognised by the international community. Hence the reputational effect of

compliance will be greater for the belligerent.

Model 7: Sequential game for constrained belligerent

This increased gain for compliance has no immediate impact on the equilibrium, but same as for

Model 6, the agreement in place will affect the expectations of the opponent’s behaviour. In all

of the models it is in fact the agreement that puts a credible commitment in place, which in turn

affects the expectation of the other’s response that exerts the pull to compliance.